01_neural_network_regression_in_tensorflow.ipynb

{
  "nbformat": 4,
  "nbformat_minor": 0,
  "metadata": {
    "colab": {
      "provenance": [],
      "collapsed_sections": [],
      "toc_visible": true,
      "authorship_tag": "ABX9TyOtNf+N/iiiARMRBYjkhgPZ",
      "include_colab_link": true
    },
    "kernelspec": {
      "name": "python3",
      "display_name": "Python 3"
    },
    "language_info": {
      "name": "python"
    }
  },
  "cells": [
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "view-in-github",
        "colab_type": "text"
      },
      "source": [
        "<a href=\"https://colab.research.google.com/github/SaketMunda/tensorflow-fundamentals/blob/master/01_neural_network_regression_in_tensorflow.ipynb\" target=\"_parent\"><img src=\"https://colab.research.google.com/assets/colab-badge.svg\" alt=\"Open In Colab\"/></a>"
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "# 01. Neural Network Regression with TensorFlow\n",
        "\n",
        "In this notebook, we're going to set the foundations for how you can take a sample of inputs(this is our data), build a neural network to discover patterns in those inputs and then make a prediction(in form of a number) based on those inputs."
      ],
      "metadata": {
        "id": "W6f0UacUZj40"
      }
    },
    {
      "cell_type": "markdown",
      "source": [
        "## Typical Architecture of Regression Neural Network\n",
        "\n",
        "The word *typical* is on purpose.\n",
        "\n",
        "Why?\n",
        "\n",
        "Because there are many different ways (actually, there's almost an infinite number of ways) to write neural network.\n",
        "\n",
        "But the following is a generic setup for ingesting a collection of numbers, finding patterns in them and then outputing some kind of target number.\n",
        "\n",
        "\n",
        "| **Hyperparameter** | **Typical value** |\n",
        "| --- | --- |\n",
        "| Input layer shape | Same shape as number of features (e.g. 3 for # bedrooms, # bathrooms, # car spaces in housing price prediction) |\n",
        "| Hidden layer(s) | Problem specific, minimum = 1, maximum = unlimited |\n",
        "| Neurons per hidden layer | Problem specific, generally 10 to 100 |\n",
        "| Output layer shape | Same shape as desired prediction shape (e.g. 1 for house price) |\n",
        "| Hidden activation | Usually [ReLU](https://www.kaggle.com/dansbecker/rectified-linear-units-relu-in-deep-learning) (rectified linear unit) |\n",
        "| Output activation | None, ReLU, logistic/tanh |\n",
        "| Loss function | [MSE](https://en.wikipedia.org/wiki/Mean_squared_error) (mean square error) or [MAE](https://en.wikipedia.org/wiki/Mean_absolute_error) (mean absolute error)/Huber (combination of MAE/MSE) if outliers |\n",
        "| Optimizer | [SGD](https://www.tensorflow.org/api_docs/python/tf/keras/optimizers/SGD) (stochastic gradient descent), [Adam](https://www.tensorflow.org/api_docs/python/tf/keras/optimizers/Adam) |\n",
        "\n",
        "***Table 1:*** *Typical architecture of a regression network.* ***Source:*** *Adapted from page 293 of [Hands-On Machine Learning with Scikit-Learn, Keras & TensorFlow Book by Aurélien Géron](https://www.oreilly.com/library/view/hands-on-machine-learning/9781492032632/)*\n"
      ],
      "metadata": {
        "id": "2yMovR5Ei1Yr"
      }
    },
    {
      "cell_type": "markdown",
      "source": [
        "> 🔑 **Note**: A **hyperparameter** in machine learning is something a data analyst or developer can set themselves, where as a **parameter** usually describes something a model learns on its own (a value not explicitly set by an analyst)."
      ],
      "metadata": {
        "id": "-m6-iIs7j8aw"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "# Import the tensorFlow\n",
        "import tensorflow as tf\n",
        "print(tf.__version__)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "r5DzVSjBk5p5",
        "outputId": "9d593a84-f6ff-446d-b262-22814e01e50f"
      },
      "execution_count": 1,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "2.9.2\n"
          ]
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "## Creating data to view and fit\n",
        "\n",
        "Since we're working on a regression problem (predicting a number), let's create some linear data (a straight line) to model."
      ],
      "metadata": {
        "id": "LsFdN0L2m5Kz"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "import numpy as np\n",
        "import matplotlib.pyplot as plt\n",
        "\n",
        "# Create features\n",
        "X = np.array([-7.0, -4.0, -1.0, 2.0, 5.0, 8.0, 11.0, 14.0])\n",
        "\n",
        "# Create labels\n",
        "y = np.array([-3.0, 0.0, 3.0, 6.0, 9.0, 12.0, 15.0, 18.0])\n",
        "\n",
        "# Visualise it\n",
        "plt.scatter(X,y);"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 265
        },
        "id": "12LMC9ADnLj3",
        "outputId": "d721741b-9f2e-4d5e-b229-1879a443d27a"
      },
      "execution_count": 2,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "Before we do any modelling, can we calculate the pattern between X and y?\n",
        "\n",
        "For example, say I asked you, based on this data what the y value would be if X was 17.0 ?\n",
        "\n",
        "or how about if X was -10.0 ?\n",
        "\n",
        "This kind of pattern discovery is the essence of what we'll be building neural networks to do for us."
      ],
      "metadata": {
        "id": "GfE5U1OAnRe0"
      }
    },
    {
      "cell_type": "markdown",
      "source": [
        "## Regression input shapes and output shapes\n",
        "\n",
        "One of the most important concepts when working with neural networks are the input and output shapes.\n",
        "\n",
        "The **input shape** is the shape of your data that goes into the model.\n",
        "The **output shape** is the shape of your data you want to come out of your model.\n",
        "\n",
        "These will differ depending on the problem you're working on.\n",
        "\n",
        "Neural networks accept numbers and output numbers. These numbers are typically represented as tensors(or arrays).\n",
        "\n",
        "Before, we created data using Numpy arrays, but we could do the same with tensors."
      ],
      "metadata": {
        "id": "Nhwn395rpYZ4"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "# Example input and output shapes of a regression model\n",
        "house_info = tf.constant([\"bedroom\",\"bathroom\", \"garage\"])\n",
        "house_price = tf.constant([939700])\n",
        "house_info, house_price"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "h5ifuXF5uz49",
        "outputId": "a5788f5b-719c-41ac-c8c1-5daccd145f81"
      },
      "execution_count": 3,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "(<tf.Tensor: shape=(3,), dtype=string, numpy=array([b'bedroom', b'bathroom', b'garage'], dtype=object)>,\n",
              " <tf.Tensor: shape=(1,), dtype=int32, numpy=array([939700], dtype=int32)>)"
            ]
          },
          "metadata": {},
          "execution_count": 3
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "house_info.shape, house_price.shape"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "go-zW8dQvEzF",
        "outputId": "925311db-803b-4840-a322-6aec1ba103d7"
      },
      "execution_count": 4,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "(TensorShape([3]), TensorShape([1]))"
            ]
          },
          "metadata": {},
          "execution_count": 4
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "It says input shape as 3, and output shape is 1"
      ],
      "metadata": {
        "id": "YAOQb8jdvowI"
      }
    },
    {
      "cell_type": "markdown",
      "source": [
        "\n",
        "![example of input and output shapes for a housing price prediction problem](https://raw.githubusercontent.com/mrdbourke/tensorflow-deep-learning/main/images/01-input-and-output-shapes-housing-prices.png)\n",
        "\n",
        "*If we were working on building a machine learning algorithm for predicting housing prices, our inputs may be number of bedrooms, number of bathrooms and number of garages, giving you an input shape of 3 (3 different features). And since we are trying to predict the price of the house, our output shape would be 1.*"
      ],
      "metadata": {
        "id": "1r52eIn2vtWI"
      }
    },
    {
      "cell_type": "markdown",
      "source": [
        "## Steps in modelling with TensorFlow\n",
        "\n",
        "Now we know what data we have as well as the input and output shapes, let's see how we'd build a neural network to model it.\n",
        "\n",
        "In TensorFlow, there are typically 3 fundamental steps to creating and training a model.\n",
        "\n",
        "1. **Creating a model**: piece together the layers of a neural network yourself (using Functional or Sequential API) or import a previously build model (known as transfer learning).\n",
        "2. **Compiling a model**: defining how a model performance should be measured(loss/metrics) as well as defining how it should improve(optimizer)\n",
        "3. **Fitting a model**: letting the model try to find patterns in the data(how does x get to y)\n",
        "\n",
        "\n",
        "Let's see using the Keras Sequential API to build a model for our regression problem."
      ],
      "metadata": {
        "id": "2V3Y_GjZw4FR"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "# set the random see\n",
        "tf.random.set_seed(42)\n",
        "\n",
        "# create a model using the Sequential API\n",
        "model = tf.keras.Sequential([\n",
        "    tf.keras.layers.Dense(1)\n",
        "])\n",
        "\n",
        "# Dense means, fully connected layer, it means all the neurons are connected with all the neurons of the next layer\n",
        "\n",
        "# compile the model\n",
        "\n",
        "model.compile(loss= tf.keras.losses.mae, # Mean absolute error\n",
        "              optimizer = tf.keras.optimizers.SGD(), # Stochastic gradient descent\n",
        "              metrics = [\"mae\"])\n",
        "\n",
        "# Fit the model\n",
        "model.fit(tf.expand_dims(X, axis = -1), y, epochs = 5)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "JIK-rK-PyYlJ",
        "outputId": "6543ba7d-f544-45e4-b189-f0a6dee3c720"
      },
      "execution_count": 5,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Epoch 1/5\n",
            "1/1 [==============================] - 1s 1s/step - loss: 5.5048 - mae: 5.5048\n",
            "Epoch 2/5\n",
            "1/1 [==============================] - 0s 11ms/step - loss: 5.3723 - mae: 5.3723\n",
            "Epoch 3/5\n",
            "1/1 [==============================] - 0s 30ms/step - loss: 5.2398 - mae: 5.2398\n",
            "Epoch 4/5\n",
            "1/1 [==============================] - 0s 22ms/step - loss: 5.1073 - mae: 5.1073\n",
            "Epoch 5/5\n",
            "1/1 [==============================] - 0s 7ms/step - loss: 4.9748 - mae: 4.9748\n"
          ]
        },
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "<keras.callbacks.History at 0x7fe5bd6cb1d0>"
            ]
          },
          "metadata": {},
          "execution_count": 5
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "We've just trained a model to figure out the patterns between X and y\n"
      ],
      "metadata": {
        "id": "IVs1iOFgzZqW"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "# Check out X and y\n",
        "X, y"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "CdeuHOZSzobM",
        "outputId": "87268a6e-8520-4d50-be03-ba8b433916c8"
      },
      "execution_count": 6,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "(array([-7., -4., -1.,  2.,  5.,  8., 11., 14.]),\n",
              " array([-3.,  0.,  3.,  6.,  9., 12., 15., 18.]))"
            ]
          },
          "metadata": {},
          "execution_count": 6
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "What do you think the outcome should be if we passed our model an x value of 17.0 ?\n"
      ],
      "metadata": {
        "id": "_B_iEtluzqSp"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "# Make a prediction with the model\n",
        "model.predict([17.0])"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "gfEHXWfnzzK8",
        "outputId": "44a75ebf-62d1-445d-b121-eb4678752cee"
      },
      "execution_count": 7,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "1/1 [==============================] - 0s 278ms/step\n"
          ]
        },
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "array([[12.716021]], dtype=float32)"
            ]
          },
          "metadata": {},
          "execution_count": 7
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "12.71 + 4.97 # adding with our model's loss"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "YRDgEX-R0YhF",
        "outputId": "f863c8ef-e305-4b29-bb65-86094c38b6b3"
      },
      "execution_count": 8,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "17.68"
            ]
          },
          "metadata": {},
          "execution_count": 8
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "It doesn't went well.. it should've output something close to 21.0\n",
        "\n",
        "> 🤔 **What's Keras ?** : I thought we were working with TensorFLow but everytime we write TensorFlow code, `keras` comes after `tf` (e.g `tf.keras.layers.Dense()`) ?\n",
        "\n",
        "Before TensorFlow 2.0+, Keras was an API designed to be able to build deep learning models with ease. Since TensorFlow 2.0+, its functionality has been tightly integrated withing the TensorFlow Library."
      ],
      "metadata": {
        "id": "UW63yVq00dqF"
      }
    },
    {
      "cell_type": "markdown",
      "source": [
        "## Improving a model\n",
        "\n",
        "To improve our model, we alter almost every part of 3 steps we went through before.\n",
        "\n",
        "1. **Creating a model**: here we might add some more layers, increase the number of hidden units (also called as neurons) within each layer, change the activation functions of each layer.\n",
        "2. **Compiling a model**: we might want to choose different optimization function or perhaps change the **learning rate** of the optimization function.\n",
        "3. **Fitting a model**: perhaps we might want to train on more data or we could fit a model for more **epochs** (leave it training for longer)\n",
        "\n",
        "![improve a model](https://raw.githubusercontent.com/mrdbourke/tensorflow-deep-learning/main/images/02-improving-a-model-from-model-perspective.png)\n",
        "\n",
        "*There are many different ways to potentially improve a neural network. Some of the most common include: increasing the number of layers (making the network deeper), increasing the number of hidden units (making the network wider) and changing the learning rate. Because these values are all human-changeable, they're referred to as hyperparameters and the practice of trying to find the best hyperparameters is referred to as hyperparameter tuning.*\n",
        "\n",
        "\n",
        "\n",
        "For now, let's keep it simple, all we'll do is to train our model for longer (everything else will stay the same)"
      ],
      "metadata": {
        "id": "T0K_Iedy1kM1"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "# set random seed\n",
        "tf.random.set_seed(42)\n",
        "\n",
        "# create a model (same as above)\n",
        "model = tf.keras.Sequential([\n",
        "    tf.keras.layers.Dense(1)\n",
        "])\n",
        "\n",
        "# compile a model\n",
        "model.compile(loss=tf.keras.losses.mae,\n",
        "              optimizer=tf.keras.optimizers.SGD(),\n",
        "              metrics=[\"mae\"])\n",
        "\n",
        "# fitting a model\n",
        "model.fit(tf.expand_dims(X, axis = -1),y,epochs = 100)"
      ],
      "metadata": {
        "id": "coYGyqg41m8A",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "30489aa1-f25c-4981-e804-c6ca57733037"
      },
      "execution_count": 9,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Epoch 1/100\n",
            "1/1 [==============================] - 1s 695ms/step - loss: 5.5048 - mae: 5.5048\n",
            "Epoch 2/100\n",
            "1/1 [==============================] - 0s 13ms/step - loss: 5.3723 - mae: 5.3723\n",
            "Epoch 3/100\n",
            "1/1 [==============================] - 0s 21ms/step - loss: 5.2398 - mae: 5.2398\n",
            "Epoch 4/100\n",
            "1/1 [==============================] - 0s 14ms/step - loss: 5.1073 - mae: 5.1073\n",
            "Epoch 5/100\n",
            "1/1 [==============================] - 0s 9ms/step - loss: 4.9748 - mae: 4.9748\n",
            "Epoch 6/100\n",
            "1/1 [==============================] - 0s 18ms/step - loss: 4.8423 - mae: 4.8423\n",
            "Epoch 7/100\n",
            "1/1 [==============================] - 0s 12ms/step - loss: 4.7098 - mae: 4.7098\n",
            "Epoch 8/100\n",
            "1/1 [==============================] - 0s 15ms/step - loss: 4.5773 - mae: 4.5773\n",
            "Epoch 9/100\n",
            "1/1 [==============================] - 0s 14ms/step - loss: 4.4448 - mae: 4.4448\n",
            "Epoch 10/100\n",
            "1/1 [==============================] - 0s 11ms/step - loss: 4.3123 - mae: 4.3123\n",
            "Epoch 11/100\n",
            "1/1 [==============================] - 0s 12ms/step - loss: 4.1798 - mae: 4.1798\n",
            "Epoch 12/100\n",
            "1/1 [==============================] - 0s 15ms/step - loss: 4.0473 - mae: 4.0473\n",
            "Epoch 13/100\n",
            "1/1 [==============================] - 0s 10ms/step - loss: 3.9148 - mae: 3.9148\n",
            "Epoch 14/100\n",
            "1/1 [==============================] - 0s 10ms/step - loss: 3.7823 - mae: 3.7823\n",
            "Epoch 15/100\n",
            "1/1 [==============================] - 0s 10ms/step - loss: 3.6498 - mae: 3.6498\n",
            "Epoch 16/100\n",
            "1/1 [==============================] - 0s 12ms/step - loss: 3.5173 - mae: 3.5173\n",
            "Epoch 17/100\n",
            "1/1 [==============================] - 0s 17ms/step - loss: 3.3848 - mae: 3.3848\n",
            "Epoch 18/100\n",
            "1/1 [==============================] - 0s 16ms/step - loss: 3.2523 - mae: 3.2523\n",
            "Epoch 19/100\n",
            "1/1 [==============================] - 0s 27ms/step - loss: 3.1198 - mae: 3.1198\n",
            "Epoch 20/100\n",
            "1/1 [==============================] - 0s 15ms/step - loss: 2.9873 - mae: 2.9873\n",
            "Epoch 21/100\n",
            "1/1 [==============================] - 0s 17ms/step - loss: 2.8548 - mae: 2.8548\n",
            "Epoch 22/100\n",
            "1/1 [==============================] - 0s 19ms/step - loss: 2.8425 - mae: 2.8425\n",
            "Epoch 23/100\n",
            "1/1 [==============================] - 0s 16ms/step - loss: 2.8369 - mae: 2.8369\n",
            "Epoch 24/100\n",
            "1/1 [==============================] - 0s 13ms/step - loss: 2.8312 - mae: 2.8312\n",
            "Epoch 25/100\n",
            "1/1 [==============================] - 0s 30ms/step - loss: 2.8256 - mae: 2.8256\n",
            "Epoch 26/100\n",
            "1/1 [==============================] - 0s 17ms/step - loss: 2.8200 - mae: 2.8200\n",
            "Epoch 27/100\n",
            "1/1 [==============================] - 0s 10ms/step - loss: 2.8144 - mae: 2.8144\n",
            "Epoch 28/100\n",
            "1/1 [==============================] - 0s 18ms/step - loss: 2.8087 - mae: 2.8087\n",
            "Epoch 29/100\n",
            "1/1 [==============================] - 0s 34ms/step - loss: 2.8031 - mae: 2.8031\n",
            "Epoch 30/100\n",
            "1/1 [==============================] - 0s 8ms/step - loss: 2.7975 - mae: 2.7975\n",
            "Epoch 31/100\n",
            "1/1 [==============================] - 0s 8ms/step - loss: 2.7919 - mae: 2.7919\n",
            "Epoch 32/100\n",
            "1/1 [==============================] - 0s 7ms/step - loss: 2.7863 - mae: 2.7863\n",
            "Epoch 33/100\n",
            "1/1 [==============================] - 0s 19ms/step - loss: 2.7806 - mae: 2.7806\n",
            "Epoch 34/100\n",
            "1/1 [==============================] - 0s 12ms/step - loss: 2.7750 - mae: 2.7750\n",
            "Epoch 35/100\n",
            "1/1 [==============================] - 0s 19ms/step - loss: 2.7694 - mae: 2.7694\n",
            "Epoch 36/100\n",
            "1/1 [==============================] - 0s 11ms/step - loss: 2.7638 - mae: 2.7638\n",
            "Epoch 37/100\n",
            "1/1 [==============================] - 0s 13ms/step - loss: 2.7581 - mae: 2.7581\n",
            "Epoch 38/100\n",
            "1/1 [==============================] - 0s 14ms/step - loss: 2.7525 - mae: 2.7525\n",
            "Epoch 39/100\n",
            "1/1 [==============================] - 0s 15ms/step - loss: 2.7469 - mae: 2.7469\n",
            "Epoch 40/100\n",
            "1/1 [==============================] - 0s 14ms/step - loss: 2.7413 - mae: 2.7413\n",
            "Epoch 41/100\n",
            "1/1 [==============================] - 0s 18ms/step - loss: 2.7356 - mae: 2.7356\n",
            "Epoch 42/100\n",
            "1/1 [==============================] - 0s 18ms/step - loss: 2.7300 - mae: 2.7300\n",
            "Epoch 43/100\n",
            "1/1 [==============================] - 0s 17ms/step - loss: 2.7244 - mae: 2.7244\n",
            "Epoch 44/100\n",
            "1/1 [==============================] - 0s 15ms/step - loss: 2.7188 - mae: 2.7188\n",
            "Epoch 45/100\n",
            "1/1 [==============================] - 0s 10ms/step - loss: 2.7131 - mae: 2.7131\n",
            "Epoch 46/100\n",
            "1/1 [==============================] - 0s 10ms/step - loss: 2.7075 - mae: 2.7075\n",
            "Epoch 47/100\n",
            "1/1 [==============================] - 0s 13ms/step - loss: 2.7019 - mae: 2.7019\n",
            "Epoch 48/100\n",
            "1/1 [==============================] - 0s 10ms/step - loss: 2.6963 - mae: 2.6963\n",
            "Epoch 49/100\n",
            "1/1 [==============================] - 0s 8ms/step - loss: 2.6906 - mae: 2.6906\n",
            "Epoch 50/100\n",
            "1/1 [==============================] - 0s 9ms/step - loss: 2.6850 - mae: 2.6850\n",
            "Epoch 51/100\n",
            "1/1 [==============================] - 0s 15ms/step - loss: 2.6794 - mae: 2.6794\n",
            "Epoch 52/100\n",
            "1/1 [==============================] - 0s 16ms/step - loss: 2.6738 - mae: 2.6738\n",
            "Epoch 53/100\n",
            "1/1 [==============================] - 0s 11ms/step - loss: 2.6681 - mae: 2.6681\n",
            "Epoch 54/100\n",
            "1/1 [==============================] - 0s 7ms/step - loss: 2.6625 - mae: 2.6625\n",
            "Epoch 55/100\n",
            "1/1 [==============================] - 0s 8ms/step - loss: 2.6569 - mae: 2.6569\n",
            "Epoch 56/100\n",
            "1/1 [==============================] - 0s 8ms/step - loss: 2.6513 - mae: 2.6513\n",
            "Epoch 57/100\n",
            "1/1 [==============================] - 0s 7ms/step - loss: 2.6456 - mae: 2.6456\n",
            "Epoch 58/100\n",
            "1/1 [==============================] - 0s 12ms/step - loss: 2.6400 - mae: 2.6400\n",
            "Epoch 59/100\n",
            "1/1 [==============================] - 0s 11ms/step - loss: 2.6344 - mae: 2.6344\n",
            "Epoch 60/100\n",
            "1/1 [==============================] - 0s 12ms/step - loss: 2.6288 - mae: 2.6288\n",
            "Epoch 61/100\n",
            "1/1 [==============================] - 0s 13ms/step - loss: 2.6231 - mae: 2.6231\n",
            "Epoch 62/100\n",
            "1/1 [==============================] - 0s 16ms/step - loss: 2.6175 - mae: 2.6175\n",
            "Epoch 63/100\n",
            "1/1 [==============================] - 0s 35ms/step - loss: 2.6119 - mae: 2.6119\n",
            "Epoch 64/100\n",
            "1/1 [==============================] - 0s 11ms/step - loss: 2.6063 - mae: 2.6063\n",
            "Epoch 65/100\n",
            "1/1 [==============================] - 0s 14ms/step - loss: 2.6006 - mae: 2.6006\n",
            "Epoch 66/100\n",
            "1/1 [==============================] - 0s 13ms/step - loss: 2.5950 - mae: 2.5950\n",
            "Epoch 67/100\n",
            "1/1 [==============================] - 0s 17ms/step - loss: 2.5894 - mae: 2.5894\n",
            "Epoch 68/100\n",
            "1/1 [==============================] - 0s 19ms/step - loss: 2.5838 - mae: 2.5838\n",
            "Epoch 69/100\n",
            "1/1 [==============================] - 0s 14ms/step - loss: 2.5781 - mae: 2.5781\n",
            "Epoch 70/100\n",
            "1/1 [==============================] - 0s 18ms/step - loss: 2.5725 - mae: 2.5725\n",
            "Epoch 71/100\n",
            "1/1 [==============================] - 0s 15ms/step - loss: 2.5669 - mae: 2.5669\n",
            "Epoch 72/100\n",
            "1/1 [==============================] - 0s 19ms/step - loss: 2.5613 - mae: 2.5613\n",
            "Epoch 73/100\n",
            "1/1 [==============================] - 0s 16ms/step - loss: 2.5556 - mae: 2.5556\n",
            "Epoch 74/100\n",
            "1/1 [==============================] - 0s 22ms/step - loss: 2.5500 - mae: 2.5500\n",
            "Epoch 75/100\n",
            "1/1 [==============================] - 0s 34ms/step - loss: 2.5444 - mae: 2.5444\n",
            "Epoch 76/100\n",
            "1/1 [==============================] - 0s 17ms/step - loss: 2.5388 - mae: 2.5388\n",
            "Epoch 77/100\n",
            "1/1 [==============================] - 0s 20ms/step - loss: 2.5331 - mae: 2.5331\n",
            "Epoch 78/100\n",
            "1/1 [==============================] - 0s 12ms/step - loss: 2.5275 - mae: 2.5275\n",
            "Epoch 79/100\n",
            "1/1 [==============================] - 0s 8ms/step - loss: 2.5219 - mae: 2.5219\n",
            "Epoch 80/100\n",
            "1/1 [==============================] - 0s 11ms/step - loss: 2.5164 - mae: 2.5164\n",
            "Epoch 81/100\n",
            "1/1 [==============================] - 0s 9ms/step - loss: 2.5125 - mae: 2.5125\n",
            "Epoch 82/100\n",
            "1/1 [==============================] - 0s 8ms/step - loss: 2.5069 - mae: 2.5069\n",
            "Epoch 83/100\n",
            "1/1 [==============================] - 0s 11ms/step - loss: 2.5013 - mae: 2.5013\n",
            "Epoch 84/100\n",
            "1/1 [==============================] - 0s 16ms/step - loss: 2.4956 - mae: 2.4956\n",
            "Epoch 85/100\n",
            "1/1 [==============================] - 0s 11ms/step - loss: 2.4900 - mae: 2.4900\n",
            "Epoch 86/100\n",
            "1/1 [==============================] - 0s 13ms/step - loss: 2.4844 - mae: 2.4844\n",
            "Epoch 87/100\n",
            "1/1 [==============================] - 0s 9ms/step - loss: 2.4788 - mae: 2.4788\n",
            "Epoch 88/100\n",
            "1/1 [==============================] - 0s 24ms/step - loss: 2.4731 - mae: 2.4731\n",
            "Epoch 89/100\n",
            "1/1 [==============================] - 0s 10ms/step - loss: 2.4675 - mae: 2.4675\n",
            "Epoch 90/100\n",
            "1/1 [==============================] - 0s 15ms/step - loss: 2.4619 - mae: 2.4619\n",
            "Epoch 91/100\n",
            "1/1 [==============================] - 0s 8ms/step - loss: 2.4563 - mae: 2.4563\n",
            "Epoch 92/100\n",
            "1/1 [==============================] - 0s 15ms/step - loss: 2.4506 - mae: 2.4506\n",
            "Epoch 93/100\n",
            "1/1 [==============================] - 0s 27ms/step - loss: 2.4450 - mae: 2.4450\n",
            "Epoch 94/100\n",
            "1/1 [==============================] - 0s 23ms/step - loss: 2.4394 - mae: 2.4394\n",
            "Epoch 95/100\n",
            "1/1 [==============================] - 0s 26ms/step - loss: 2.4338 - mae: 2.4338\n",
            "Epoch 96/100\n",
            "1/1 [==============================] - 0s 13ms/step - loss: 2.4281 - mae: 2.4281\n",
            "Epoch 97/100\n",
            "1/1 [==============================] - 0s 11ms/step - loss: 2.4225 - mae: 2.4225\n",
            "Epoch 98/100\n",
            "1/1 [==============================] - 0s 8ms/step - loss: 2.4169 - mae: 2.4169\n",
            "Epoch 99/100\n",
            "1/1 [==============================] - 0s 14ms/step - loss: 2.4113 - mae: 2.4113\n",
            "Epoch 100/100\n",
            "1/1 [==============================] - 0s 8ms/step - loss: 2.4056 - mae: 2.4056\n"
          ]
        },
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "<keras.callbacks.History at 0x7fe5b88f7990>"
            ]
          },
          "metadata": {},
          "execution_count": 9
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "If you notice, we have minimised the loss just below 2.5 which was almost 5 when we were fitting the model with 5 `epochs`"
      ],
      "metadata": {
        "id": "51wl8sxvKe_4"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "# let's see X & y\n",
        "X, y"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "AhLJUyasLBXc",
        "outputId": "de6bd511-9d26-46f3-d5d6-40306400d6b7"
      },
      "execution_count": 10,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "(array([-7., -4., -1.,  2.,  5.,  8., 11., 14.]),\n",
              " array([-3.,  0.,  3.,  6.,  9., 12., 15., 18.]))"
            ]
          },
          "metadata": {},
          "execution_count": 10
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "# make prediction with improved model\n",
        "model.predict([17.0])"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "7mfCXNrPLFGM",
        "outputId": "fc26e863-1867-47cb-e31f-ea32b5c0f6de"
      },
      "execution_count": 11,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "1/1 [==============================] - 0s 212ms/step\n"
          ]
        },
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "array([[22.518517]], dtype=float32)"
            ]
          },
          "metadata": {},
          "execution_count": 11
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "It should predict around 21.0, so our model is almost closer to the correct predictions.\n",
        "\n",
        "Let's try tuning other hyperparameters,\n",
        "\n",
        "This time, we will choose different optimizer function"
      ],
      "metadata": {
        "id": "vUXd5v3dLOIy"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "# again do the same process,\n",
        "\n",
        "# set random seed\n",
        "tf.random.set_seed(42)\n",
        "\n",
        "# create a model\n",
        "model = tf.keras.Sequential([\n",
        "    tf.keras.layers.Dense(1)\n",
        "])\n",
        "\n",
        "# compile a model\n",
        "model.compile(loss=tf.keras.losses.mae,\n",
        "              optimizer=tf.keras.optimizers.Adam(learning_rate=0.0001), # lr stands for learning rate\n",
        "              metrics=[\"mae\"])\n",
        "\n",
        "# fit a model\n",
        "model.fit(tf.expand_dims(X, axis = -1), y, epochs = 100)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "nx9GPFbwLaGL",
        "outputId": "c59b7e98-a084-4443-e5b9-f09552ec2774"
      },
      "execution_count": 12,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Epoch 1/100\n",
            "1/1 [==============================] - 1s 690ms/step - loss: 5.5048 - mae: 5.5048\n",
            "Epoch 2/100\n",
            "1/1 [==============================] - 0s 21ms/step - loss: 5.5043 - mae: 5.5043\n",
            "Epoch 3/100\n",
            "1/1 [==============================] - 0s 29ms/step - loss: 5.5039 - mae: 5.5039\n",
            "Epoch 4/100\n",
            "1/1 [==============================] - 0s 17ms/step - loss: 5.5034 - mae: 5.5034\n",
            "Epoch 5/100\n",
            "1/1 [==============================] - 0s 12ms/step - loss: 5.5030 - mae: 5.5030\n",
            "Epoch 6/100\n",
            "1/1 [==============================] - 0s 11ms/step - loss: 5.5025 - mae: 5.5025\n",
            "Epoch 7/100\n",
            "1/1 [==============================] - 0s 8ms/step - loss: 5.5021 - mae: 5.5021\n",
            "Epoch 8/100\n",
            "1/1 [==============================] - 0s 7ms/step - loss: 5.5016 - mae: 5.5016\n",
            "Epoch 9/100\n",
            "1/1 [==============================] - 0s 7ms/step - loss: 5.5012 - mae: 5.5012\n",
            "Epoch 10/100\n",
            "1/1 [==============================] - 0s 10ms/step - loss: 5.5007 - mae: 5.5007\n",
            "Epoch 11/100\n",
            "1/1 [==============================] - 0s 9ms/step - loss: 5.5003 - mae: 5.5003\n",
            "Epoch 12/100\n",
            "1/1 [==============================] - 0s 12ms/step - loss: 5.4998 - mae: 5.4998\n",
            "Epoch 13/100\n",
            "1/1 [==============================] - 0s 10ms/step - loss: 5.4994 - mae: 5.4994\n",
            "Epoch 14/100\n",
            "1/1 [==============================] - 0s 16ms/step - loss: 5.4989 - mae: 5.4989\n",
            "Epoch 15/100\n",
            "1/1 [==============================] - 0s 12ms/step - loss: 5.4985 - mae: 5.4985\n",
            "Epoch 16/100\n",
            "1/1 [==============================] - 0s 9ms/step - loss: 5.4980 - mae: 5.4980\n",
            "Epoch 17/100\n",
            "1/1 [==============================] - 0s 7ms/step - loss: 5.4976 - mae: 5.4976\n",
            "Epoch 18/100\n",
            "1/1 [==============================] - 0s 9ms/step - loss: 5.4971 - mae: 5.4971\n",
            "Epoch 19/100\n",
            "1/1 [==============================] - 0s 8ms/step - loss: 5.4967 - mae: 5.4967\n",
            "Epoch 20/100\n",
            "1/1 [==============================] - 0s 8ms/step - loss: 5.4962 - mae: 5.4962\n",
            "Epoch 21/100\n",
            "1/1 [==============================] - 0s 8ms/step - loss: 5.4958 - mae: 5.4958\n",
            "Epoch 22/100\n",
            "1/1 [==============================] - 0s 9ms/step - loss: 5.4953 - mae: 5.4953\n",
            "Epoch 23/100\n",
            "1/1 [==============================] - 0s 8ms/step - loss: 5.4949 - mae: 5.4949\n",
            "Epoch 24/100\n",
            "1/1 [==============================] - 0s 8ms/step - loss: 5.4944 - mae: 5.4944\n",
            "Epoch 25/100\n",
            "1/1 [==============================] - 0s 8ms/step - loss: 5.4940 - mae: 5.4940\n",
            "Epoch 26/100\n",
            "1/1 [==============================] - 0s 9ms/step - loss: 5.4935 - mae: 5.4935\n",
            "Epoch 27/100\n",
            "1/1 [==============================] - 0s 8ms/step - loss: 5.4931 - mae: 5.4931\n",
            "Epoch 28/100\n",
            "1/1 [==============================] - 0s 9ms/step - loss: 5.4926 - mae: 5.4926\n",
            "Epoch 29/100\n",
            "1/1 [==============================] - 0s 15ms/step - loss: 5.4922 - mae: 5.4922\n",
            "Epoch 30/100\n",
            "1/1 [==============================] - 0s 14ms/step - loss: 5.4917 - mae: 5.4917\n",
            "Epoch 31/100\n",
            "1/1 [==============================] - 0s 15ms/step - loss: 5.4913 - mae: 5.4913\n",
            "Epoch 32/100\n",
            "1/1 [==============================] - 0s 13ms/step - loss: 5.4908 - mae: 5.4908\n",
            "Epoch 33/100\n",
            "1/1 [==============================] - 0s 10ms/step - loss: 5.4904 - mae: 5.4904\n",
            "Epoch 34/100\n",
            "1/1 [==============================] - 0s 12ms/step - loss: 5.4899 - mae: 5.4899\n",
            "Epoch 35/100\n",
            "1/1 [==============================] - 0s 8ms/step - loss: 5.4895 - mae: 5.4895\n",
            "Epoch 36/100\n",
            "1/1 [==============================] - 0s 9ms/step - loss: 5.4890 - mae: 5.4890\n",
            "Epoch 37/100\n",
            "1/1 [==============================] - 0s 15ms/step - loss: 5.4886 - mae: 5.4886\n",
            "Epoch 38/100\n",
            "1/1 [==============================] - 0s 15ms/step - loss: 5.4881 - mae: 5.4881\n",
            "Epoch 39/100\n",
            "1/1 [==============================] - 0s 15ms/step - loss: 5.4877 - mae: 5.4877\n",
            "Epoch 40/100\n",
            "1/1 [==============================] - 0s 18ms/step - loss: 5.4872 - mae: 5.4872\n",
            "Epoch 41/100\n",
            "1/1 [==============================] - 0s 9ms/step - loss: 5.4868 - mae: 5.4868\n",
            "Epoch 42/100\n",
            "1/1 [==============================] - 0s 11ms/step - loss: 5.4863 - mae: 5.4863\n",
            "Epoch 43/100\n",
            "1/1 [==============================] - 0s 10ms/step - loss: 5.4859 - mae: 5.4859\n",
            "Epoch 44/100\n",
            "1/1 [==============================] - 0s 13ms/step - loss: 5.4854 - mae: 5.4854\n",
            "Epoch 45/100\n",
            "1/1 [==============================] - 0s 12ms/step - loss: 5.4850 - mae: 5.4850\n",
            "Epoch 46/100\n",
            "1/1 [==============================] - 0s 6ms/step - loss: 5.4845 - mae: 5.4845\n",
            "Epoch 47/100\n",
            "1/1 [==============================] - 0s 9ms/step - loss: 5.4841 - mae: 5.4841\n",
            "Epoch 48/100\n",
            "1/1 [==============================] - 0s 21ms/step - loss: 5.4836 - mae: 5.4836\n",
            "Epoch 49/100\n",
            "1/1 [==============================] - 0s 14ms/step - loss: 5.4832 - mae: 5.4832\n",
            "Epoch 50/100\n",
            "1/1 [==============================] - 0s 17ms/step - loss: 5.4827 - mae: 5.4827\n",
            "Epoch 51/100\n",
            "1/1 [==============================] - 0s 16ms/step - loss: 5.4823 - mae: 5.4823\n",
            "Epoch 52/100\n",
            "1/1 [==============================] - 0s 17ms/step - loss: 5.4818 - mae: 5.4818\n",
            "Epoch 53/100\n",
            "1/1 [==============================] - 0s 17ms/step - loss: 5.4814 - mae: 5.4814\n",
            "Epoch 54/100\n",
            "1/1 [==============================] - 0s 19ms/step - loss: 5.4809 - mae: 5.4809\n",
            "Epoch 55/100\n",
            "1/1 [==============================] - 0s 12ms/step - loss: 5.4805 - mae: 5.4805\n",
            "Epoch 56/100\n",
            "1/1 [==============================] - 0s 19ms/step - loss: 5.4800 - mae: 5.4800\n",
            "Epoch 57/100\n",
            "1/1 [==============================] - 0s 18ms/step - loss: 5.4796 - mae: 5.4796\n",
            "Epoch 58/100\n",
            "1/1 [==============================] - 0s 25ms/step - loss: 5.4791 - mae: 5.4791\n",
            "Epoch 59/100\n",
            "1/1 [==============================] - 0s 43ms/step - loss: 5.4787 - mae: 5.4787\n",
            "Epoch 60/100\n",
            "1/1 [==============================] - 0s 21ms/step - loss: 5.4782 - mae: 5.4782\n",
            "Epoch 61/100\n",
            "1/1 [==============================] - 0s 10ms/step - loss: 5.4778 - mae: 5.4778\n",
            "Epoch 62/100\n",
            "1/1 [==============================] - 0s 18ms/step - loss: 5.4773 - mae: 5.4773\n",
            "Epoch 63/100\n",
            "1/1 [==============================] - 0s 22ms/step - loss: 5.4769 - mae: 5.4769\n",
            "Epoch 64/100\n",
            "1/1 [==============================] - 0s 11ms/step - loss: 5.4764 - mae: 5.4764\n",
            "Epoch 65/100\n",
            "1/1 [==============================] - 0s 12ms/step - loss: 5.4760 - mae: 5.4760\n",
            "Epoch 66/100\n",
            "1/1 [==============================] - 0s 9ms/step - loss: 5.4755 - mae: 5.4755\n",
            "Epoch 67/100\n",
            "1/1 [==============================] - 0s 14ms/step - loss: 5.4751 - mae: 5.4751\n",
            "Epoch 68/100\n",
            "1/1 [==============================] - 0s 12ms/step - loss: 5.4746 - mae: 5.4746\n",
            "Epoch 69/100\n",
            "1/1 [==============================] - 0s 25ms/step - loss: 5.4742 - mae: 5.4742\n",
            "Epoch 70/100\n",
            "1/1 [==============================] - 0s 9ms/step - loss: 5.4737 - mae: 5.4737\n",
            "Epoch 71/100\n",
            "1/1 [==============================] - 0s 10ms/step - loss: 5.4733 - mae: 5.4733\n",
            "Epoch 72/100\n",
            "1/1 [==============================] - 0s 7ms/step - loss: 5.4728 - mae: 5.4728\n",
            "Epoch 73/100\n",
            "1/1 [==============================] - 0s 7ms/step - loss: 5.4724 - mae: 5.4724\n",
            "Epoch 74/100\n",
            "1/1 [==============================] - 0s 8ms/step - loss: 5.4719 - mae: 5.4719\n",
            "Epoch 75/100\n",
            "1/1 [==============================] - 0s 12ms/step - loss: 5.4715 - mae: 5.4715\n",
            "Epoch 76/100\n",
            "1/1 [==============================] - 0s 10ms/step - loss: 5.4710 - mae: 5.4710\n",
            "Epoch 77/100\n",
            "1/1 [==============================] - 0s 13ms/step - loss: 5.4706 - mae: 5.4706\n",
            "Epoch 78/100\n",
            "1/1 [==============================] - 0s 17ms/step - loss: 5.4701 - mae: 5.4701\n",
            "Epoch 79/100\n",
            "1/1 [==============================] - 0s 9ms/step - loss: 5.4697 - mae: 5.4697\n",
            "Epoch 80/100\n",
            "1/1 [==============================] - 0s 22ms/step - loss: 5.4692 - mae: 5.4692\n",
            "Epoch 81/100\n",
            "1/1 [==============================] - 0s 20ms/step - loss: 5.4688 - mae: 5.4688\n",
            "Epoch 82/100\n",
            "1/1 [==============================] - 0s 15ms/step - loss: 5.4683 - mae: 5.4683\n",
            "Epoch 83/100\n",
            "1/1 [==============================] - 0s 12ms/step - loss: 5.4679 - mae: 5.4679\n",
            "Epoch 84/100\n",
            "1/1 [==============================] - 0s 15ms/step - loss: 5.4674 - mae: 5.4674\n",
            "Epoch 85/100\n",
            "1/1 [==============================] - 0s 16ms/step - loss: 5.4670 - mae: 5.4670\n",
            "Epoch 86/100\n",
            "1/1 [==============================] - 0s 14ms/step - loss: 5.4665 - mae: 5.4665\n",
            "Epoch 87/100\n",
            "1/1 [==============================] - 0s 27ms/step - loss: 5.4661 - mae: 5.4661\n",
            "Epoch 88/100\n",
            "1/1 [==============================] - 0s 9ms/step - loss: 5.4656 - mae: 5.4656\n",
            "Epoch 89/100\n",
            "1/1 [==============================] - 0s 18ms/step - loss: 5.4652 - mae: 5.4652\n",
            "Epoch 90/100\n",
            "1/1 [==============================] - 0s 8ms/step - loss: 5.4647 - mae: 5.4647\n",
            "Epoch 91/100\n",
            "1/1 [==============================] - 0s 26ms/step - loss: 5.4643 - mae: 5.4643\n",
            "Epoch 92/100\n",
            "1/1 [==============================] - 0s 20ms/step - loss: 5.4638 - mae: 5.4638\n",
            "Epoch 93/100\n",
            "1/1 [==============================] - 0s 17ms/step - loss: 5.4634 - mae: 5.4634\n",
            "Epoch 94/100\n",
            "1/1 [==============================] - 0s 12ms/step - loss: 5.4629 - mae: 5.4629\n",
            "Epoch 95/100\n",
            "1/1 [==============================] - 0s 21ms/step - loss: 5.4625 - mae: 5.4625\n",
            "Epoch 96/100\n",
            "1/1 [==============================] - 0s 12ms/step - loss: 5.4620 - mae: 5.4620\n",
            "Epoch 97/100\n",
            "1/1 [==============================] - 0s 15ms/step - loss: 5.4616 - mae: 5.4616\n",
            "Epoch 98/100\n",
            "1/1 [==============================] - 0s 12ms/step - loss: 5.4611 - mae: 5.4611\n",
            "Epoch 99/100\n",
            "1/1 [==============================] - 0s 7ms/step - loss: 5.4607 - mae: 5.4607\n",
            "Epoch 100/100\n",
            "1/1 [==============================] - 0s 8ms/step - loss: 5.4602 - mae: 5.4602\n"
          ]
        },
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "<keras.callbacks.History at 0x7fe5b7fe2710>"
            ]
          },
          "metadata": {},
          "execution_count": 12
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "model.predict([17.0])"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "Goi7zpzEMyCv",
        "outputId": "a07de3d8-f9e7-4b3b-a73f-0f85e2be0966"
      },
      "execution_count": 13,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "1/1 [==============================] - 0s 114ms/step\n"
          ]
        },
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "array([[9.871047]], dtype=float32)"
            ]
          },
          "metadata": {},
          "execution_count": 13
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "It is possible that not always tweaking with hyperparameters will get closer to the correct results, like in this case, our model performs worse than previous tuning, it means `Adam()` function is not right for this model. \n",
        "\n",
        "Hence, that's idea behind tweaking and improving our model, to find out best possible hyperparameters to build a perfect model."
      ],
      "metadata": {
        "id": "3JBGFRBaNBj3"
      }
    },
    {
      "cell_type": "markdown",
      "source": [
        "## Evaluating a model\n",
        "\n",
        "A typical workflow you'll go through when building neural network is:\n",
        "\n",
        "`Build a model -> evaluate it -> build (tweak) a model -> evaluate it -> build (tweak) a model -> evaluate it...`\n",
        "\n",
        "The tweaking comes from maybe not building a model from scratch but adjusting existing one.\n",
        "\n",
        "\n",
        "When it comes to evaluation, we must remember these 3 words:\n",
        "> **Visualise, Visualise, Visualise**\n",
        "\n",
        "This is because we're probably better looking at something(doing) than we are thinking about something.\n",
        "\n",
        "It's a good idea to visualise:\n",
        "* **The data** - what data are we working with? What does it look like?\n",
        "* **The model itself** - what does the architecture look like? What are the different shapes.\n",
        "* **The training of a model** - how does a model perform while it learns ?\n",
        "* **The predictions of a model** - how do the predictions of a model line up against the ground truth (the original labels)?\n",
        "\n",
        "Let's visualising our model,\n",
        "\n",
        "But first we need to create a little bit of a bigger dataset and a new model we can use."
      ],
      "metadata": {
        "id": "OSbH_Kd0Nrff"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "# Make a bigger dataset\n",
        "X = tf.range(-100,100,4)\n",
        "X"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "VqNKZ34zbJ2K",
        "outputId": "3ade152d-a024-4104-8bbf-fd6f3fcc88d6"
      },
      "execution_count": 14,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "<tf.Tensor: shape=(50,), dtype=int32, numpy=\n",
              "array([-100,  -96,  -92,  -88,  -84,  -80,  -76,  -72,  -68,  -64,  -60,\n",
              "        -56,  -52,  -48,  -44,  -40,  -36,  -32,  -28,  -24,  -20,  -16,\n",
              "        -12,   -8,   -4,    0,    4,    8,   12,   16,   20,   24,   28,\n",
              "         32,   36,   40,   44,   48,   52,   56,   60,   64,   68,   72,\n",
              "         76,   80,   84,   88,   92,   96], dtype=int32)>"
            ]
          },
          "metadata": {},
          "execution_count": 14
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "# make labels for the dataset\n",
        "# we want our model to learn the model as y = X + 5\n",
        "y = X + 5\n",
        "y"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "n4WmhcRic3i0",
        "outputId": "53cb95ad-ec25-4103-d31a-63bf05d0292c"
      },
      "execution_count": 15,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "<tf.Tensor: shape=(50,), dtype=int32, numpy=\n",
              "array([-95, -91, -87, -83, -79, -75, -71, -67, -63, -59, -55, -51, -47,\n",
              "       -43, -39, -35, -31, -27, -23, -19, -15, -11,  -7,  -3,   1,   5,\n",
              "         9,  13,  17,  21,  25,  29,  33,  37,  41,  45,  49,  53,  57,\n",
              "        61,  65,  69,  73,  77,  81,  85,  89,  93,  97, 101], dtype=int32)>"
            ]
          },
          "metadata": {},
          "execution_count": 15
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "### Split data into training/test set\n",
        "\n",
        "Usually when dealing with real-world data, spliting is typically done right at the start of the project (the test set should always be kept separate from all other data).\n",
        "\n",
        "We want our model to learn on training data and then evaluate it on test data to get an indication of how well it generalizes to unseen examples."
      ],
      "metadata": {
        "id": "s_Z1Zuy0dA8a"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "# check the number of samples we have\n",
        "len(X)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "yP8buvVYjIXS",
        "outputId": "08097755-c3d1-4b54-baa2-5aa96a9120da"
      },
      "execution_count": 16,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "50"
            ]
          },
          "metadata": {},
          "execution_count": 16
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "# Split data into train and test sets\n",
        "X_train = X[:40]  # first 40 examples (80% of data)\n",
        "y_train = y[:40]\n",
        "\n",
        "X_test = X[40:] # last 10 examples (20% of data)\n",
        "y_test = y[40:]\n",
        "\n",
        "len(X_train), len(X_test)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "uekAQyc_jLml",
        "outputId": "f7cab7c8-db34-430a-9b9c-b0e89a112926"
      },
      "execution_count": 17,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "(40, 10)"
            ]
          },
          "metadata": {},
          "execution_count": 17
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "let's visualise it using the plots"
      ],
      "metadata": {
        "id": "F7lIXR5ojl_s"
      }
    },
    {
      "cell_type": "markdown",
      "source": [
        "### Visualise the data"
      ],
      "metadata": {
        "id": "KzL9B1GBjuFI"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "plt.figure(figsize=(10,7))\n",
        "\n",
        "# plot the training data in blue\n",
        "plt.scatter(X_train, y_train, c=\"b\", label=\"Training Data\")\n",
        "\n",
        "# Plot the test data in green\n",
        "plt.scatter(X_test, y_test, c=\"g\", label=\"Test data\")\n",
        "\n",
        "# Show the legend\n",
        "plt.legend();"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 428
        },
        "id": "Tp9DHssXjwxc",
        "outputId": "3a386ae8-8777-415e-90d3-ca07e82e99ee"
      },
      "execution_count": 18,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 720x504 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "Any time we can visualise our data, our model, our anything. It's a good idea.\n",
        "\n",
        "\n",
        "Keeping this graph in our mind, what we'll be trying is to build a model which learns the pattern in the blue dots (`X_train`) to draw the green dots (`X_test`)\n",
        "\n",
        "So, let's build our model."
      ],
      "metadata": {
        "id": "Eg02XDBckhjl"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "# Set random seed\n",
        "tf.random.set_seed(42)\n",
        "\n",
        "\n",
        "# Create a model\n",
        "model = tf.keras.Sequential([\n",
        "    tf.keras.layers.Dense(1)\n",
        "])\n",
        "\n",
        "# Compile a model\n",
        "model.compile(loss=tf.keras.losses.mae,\n",
        "              optimizer= tf.keras.optimizers.SGD(),\n",
        "              metrics=[\"mae\"])\n",
        "\n",
        "# Fit the model\n",
        "# model.fit(X_train, y_train, epochs = 100) # commented out on purpose, first we will visualise the model"
      ],
      "metadata": {
        "id": "NuEBZpRFlKTO"
      },
      "execution_count": 19,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "source": [
        "### Visualise the model\n",
        "\n",
        "After we've built a model, we might want to visualise the layers and shapes of our model, by called `summary()` on it.\n",
        "\n",
        "> 🔑 **Note**: Visualizing a model is particularly helpful when you run into input and output shapes mismatches.\n"
      ],
      "metadata": {
        "id": "axSQcA1Tl8AF"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "# let's see the summary\n",
        "model.summary()"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 305
        },
        "id": "0evj4qiMmbAF",
        "outputId": "d5dfbe29-560d-49bf-c467-ab16d1798de1"
      },
      "execution_count": 20,
      "outputs": [
        {
          "output_type": "error",
          "ename": "ValueError",
          "evalue": "ignored",
          "traceback": [
            "\u001b[0;31m---------------------------------------------------------------------------\u001b[0m",
            "\u001b[0;31mValueError\u001b[0m                                Traceback (most recent call last)",
            "\u001b[0;32m<ipython-input-20-4954ad8269f7>\u001b[0m in \u001b[0;36m<module>\u001b[0;34m\u001b[0m\n\u001b[1;32m      1\u001b[0m \u001b[0;31m# let's see the summary\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m----> 2\u001b[0;31m \u001b[0mmodel\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0msummary\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m",
            "\u001b[0;32m/usr/local/lib/python3.7/dist-packages/keras/engine/training.py\u001b[0m in \u001b[0;36msummary\u001b[0;34m(self, line_length, positions, print_fn, expand_nested, show_trainable)\u001b[0m\n\u001b[1;32m   2868\u001b[0m     \u001b[0;32mif\u001b[0m \u001b[0;32mnot\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mbuilt\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m   2869\u001b[0m       raise ValueError(\n\u001b[0;32m-> 2870\u001b[0;31m           \u001b[0;34m'This model has not yet been built. '\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m   2871\u001b[0m           \u001b[0;34m'Build the model first by calling `build()` or by calling '\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m   2872\u001b[0m           'the model on a batch of data.')\n",
            "\u001b[0;31mValueError\u001b[0m: This model has not yet been built. Build the model first by calling `build()` or by calling the model on a batch of data."
          ]
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "Since the model is not built, it's not showing us the summary.\n",
        "\n",
        "We also haven't defined the input the shape while creating a model, it we define the input shape while creating a model, it will automatically build the model.\n",
        "\n",
        "So what's the input shape of our dataset ?"
      ],
      "metadata": {
        "id": "_K-zZFzNmf__"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "# since our prediction would be predicting 1 y value from 1 X value, it means X[0] will be the shape of our input\n",
        "X[0].shape"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "Pvyxe1vDms8S",
        "outputId": "27292b41-b597-4f32-a8aa-3fd56caedafa"
      },
      "execution_count": 21,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "TensorShape([])"
            ]
          },
          "metadata": {},
          "execution_count": 21
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "In this case, our `input_shape` is `[1]`.\n",
        "\n",
        "**Note:** Usually if input_shape isn't defined, Keras tries to figure it out automatically."
      ],
      "metadata": {
        "id": "IdJRwcnknCPf"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "# building the model again\n",
        "\n",
        "# set random seed\n",
        "tf.random.set_seed(42)\n",
        "\n",
        "# Create a model\n",
        "model = tf.keras.Sequential([\n",
        "    tf.keras.layers.Dense(1, input_shape=[1]) # define the input_shape of our model\n",
        "])\n",
        "\n",
        "# Compile a model\n",
        "model.compile(loss=tf.keras.losses.mae,\n",
        "              optimizer=tf.keras.optimizers.SGD(),\n",
        "              metrics=[\"mae\"])"
      ],
      "metadata": {
        "id": "9ObI5rTQni75"
      },
      "execution_count": 22,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "model.summary()"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "m-a71NC0oAnt",
        "outputId": "dfbafa72-c526-49dc-e823-78f150649326"
      },
      "execution_count": 23,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Model: \"sequential_4\"\n",
            "_________________________________________________________________\n",
            " Layer (type)                Output Shape              Param #   \n",
            "=================================================================\n",
            " dense_4 (Dense)             (None, 1)                 2         \n",
            "                                                                 \n",
            "=================================================================\n",
            "Total params: 2\n",
            "Trainable params: 2\n",
            "Non-trainable params: 0\n",
            "_________________________________________________________________\n"
          ]
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "Calling `summary()` on our model shows us the layers it contains, the output shape and the number of parameters.\n",
        "\n",
        "* **Total Params**: Total number of parameters in the model\n",
        "* **Trainable params**: These are the parameters(patterns) the model can update as it trains.\n",
        "* **Non-Trainable params**: These parameters aren't updated during training(this is typical when you bring in the already learned patterns from other models during transfer learning).\n",
        "\n",
        "> 📖 **Resource**: For a more in-depth overview of the trainabel parameters within a layer, check out [MIT's introduction to deep learning video](https://youtu.be/njKP3FqW3Sk)\n",
        "\n",
        "> 🛠 **Exercise**: Try playing around with the number of hidden units in the `Dense` Layer (e.g `Dense(2)`, `Dense(3)`). How does this change the Total/Trainable params ? Investigate what's causing the change."
      ],
      "metadata": {
        "id": "gCvk48ZZoCYL"
      }
    },
    {
      "cell_type": "markdown",
      "source": [
        "Let's fit the model to the training data"
      ],
      "metadata": {
        "id": "covrTjq7puns"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "# Fit the model to the training data\n",
        "model.fit(X_train, y_train, epochs = 100, verbose=0) # verbose controls how much gets output"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "yEEFjt4nj3Df",
        "outputId": "dce8fd74-99c5-49d5-dbd3-1b139ee3bc3d"
      },
      "execution_count": 24,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "<keras.callbacks.History at 0x7fe5b7f71e50>"
            ]
          },
          "metadata": {},
          "execution_count": 24
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "# check the model summary\n",
        "model.summary()"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "kO_YpbV1kBYq",
        "outputId": "e224ef96-f305-47f3-8d9c-71845d380e56"
      },
      "execution_count": 25,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Model: \"sequential_4\"\n",
            "_________________________________________________________________\n",
            " Layer (type)                Output Shape              Param #   \n",
            "=================================================================\n",
            " dense_4 (Dense)             (None, 1)                 2         \n",
            "                                                                 \n",
            "=================================================================\n",
            "Total params: 2\n",
            "Trainable params: 2\n",
            "Non-trainable params: 0\n",
            "_________________________________________________________________\n"
          ]
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "Alongside summary, you can also view a 2D plot of the model using `plot_model()`"
      ],
      "metadata": {
        "id": "OgJ2pZczkFL2"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "from tensorflow.keras.utils import plot_model\n",
        "\n",
        "plot_model(model=model, show_shapes = True)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 201
        },
        "id": "gfCSl4nXkRzH",
        "outputId": "000a154b-dd7e-41d3-be9d-0aaf2d454386"
      },
      "execution_count": 26,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<IPython.core.display.Image object>"
            ]
          },
          "metadata": {},
          "execution_count": 26
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "In our case, the model we used only has an input and an output but visualizing more complicated models can be very helpful for debugging."
      ],
      "metadata": {
        "id": "1S1XSV1zkapX"
      }
    },
    {
      "cell_type": "markdown",
      "source": [
        "### Visualise the predictions\n",
        "\n",
        "Now we've got a trained model, let's visualise some prediction.\n",
        "\n",
        "To visualise the predictions, it's always a good idea to plot them against the ground truth labels.\n",
        "\n",
        "Often you'll see this in the form of `y_test` vs `y_preds` (ground truth vs predictions)\n",
        "\n",
        "First, we'll make some predictions on the test data(`X_test`), remember the model has never seen the test data."
      ],
      "metadata": {
        "id": "dxEb2yxPlGQB"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "# make predictions\n",
        "y_preds = model.predict(X_test)\n",
        "y_preds"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "6t6GrbH-lVfH",
        "outputId": "3b2b4452-e461-4313-98cf-569b7274a8c7"
      },
      "execution_count": 27,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "1/1 [==============================] - 0s 45ms/step\n"
          ]
        },
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "array([[44.477352],\n",
              "       [47.392593],\n",
              "       [50.307835],\n",
              "       [53.223076],\n",
              "       [56.138313],\n",
              "       [59.053555],\n",
              "       [61.968796],\n",
              "       [64.88403 ],\n",
              "       [67.79928 ],\n",
              "       [70.714516]], dtype=float32)"
            ]
          },
          "metadata": {},
          "execution_count": 27
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "# Ground truth label\n",
        "y_test"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "756qLaKZlhQG",
        "outputId": "84278d77-75d5-47a7-ea11-f5f94ed0990f"
      },
      "execution_count": 28,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "<tf.Tensor: shape=(10,), dtype=int32, numpy=array([ 65,  69,  73,  77,  81,  85,  89,  93,  97, 101], dtype=int32)>"
            ]
          },
          "metadata": {},
          "execution_count": 28
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "Let's plot them to make a comparision between them, for this we'll be creating a function so that we can be able to use it for later comparisons as well"
      ],
      "metadata": {
        "id": "zJKWIWiklmvJ"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "# Creating a plot_prediction function\n",
        "\n",
        "import matplotlib.pyplot as plt\n",
        "\n",
        "def plot_prediction(train_data=X_train,\n",
        "                    train_label= y_train,\n",
        "                    test_data=X_test,\n",
        "                    test_label=y_test,\n",
        "                    predictions=y_preds):\n",
        "  \"\"\"\n",
        "  This function will help to visualise the prediction using scatter plot.\n",
        "  It takes train data,train labels, test data, ground truth data and predictions and plot \n",
        "  them to visualise.\n",
        "  \"\"\"\n",
        "\n",
        "  # plot the figure\n",
        "  plt.figure(figsize=(10,7))\n",
        "\n",
        "  # plot the train data in blue\n",
        "  plt.scatter(train_data, train_label, c=\"b\", label=\"Training data\")\n",
        "\n",
        "  # plot the test data in green\n",
        "  plt.scatter(test_data, test_label, c=\"g\", label=\"Test data\")\n",
        "\n",
        "  # plot the test data vs predictions in red\n",
        "  plt.scatter(test_data, predictions, c=\"r\", label=\"Predictions\")\n",
        "\n",
        "  # show the legend\n",
        "  plt.legend();"
      ],
      "metadata": {
        "id": "ZvhbNd0GmM4q"
      },
      "execution_count": 29,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "plot_prediction()"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 428
        },
        "id": "WwZUizlEnisL",
        "outputId": "05d85f0b-2e63-4ac0-c2ed-e71c64c062f7"
      },
      "execution_count": 30,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 720x504 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "From the plot we can see our predictions aren't totally outlandish but they definitely aren't anything special either."
      ],
      "metadata": {
        "id": "R-FUfr2MoyTT"
      }
    },
    {
      "cell_type": "markdown",
      "source": [
        "### Evaluating predictions\n",
        "\n",
        "Alongside visualizations, evaluation metrics are your alternative best option for evaluating our model.\n",
        "\n",
        "Depending on the problem you're working on, different models have different evaluation metrics.\n",
        "\n",
        "Two of the main metrics used for regression problems are:\n",
        "\n",
        "* **MAE** : Mean absolute error, the mean difference between each of the predictions\n",
        "* **MSE** : Mean Squared error, the squared mean difference between of the predictions (use if larger errors are more detrimental than smaller ones)\n",
        "* **Huber** : Combination of MAE and MSE\n",
        "\n",
        "The lower each of these values, the better.\n",
        "\n",
        "We can also use `model.evaluate()` which also returns the metrics results that we mentioned while `model.compile()` as `metrics`"
      ],
      "metadata": {
        "id": "Mrr-KLEOqrNw"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "model.evaluate(X_test, y_test)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "t4JZVuAfrxMR",
        "outputId": "c18932ce-a7aa-45c7-a02c-deac1ec013fe"
      },
      "execution_count": 31,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "1/1 [==============================] - 0s 130ms/step - loss: 25.4041 - mae: 25.4041\n"
          ]
        },
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "[25.404064178466797, 25.404064178466797]"
            ]
          },
          "metadata": {},
          "execution_count": 31
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "In our case, we used MAE as the metrics, so it's showing the MAE of the model.\n",
        "\n",
        "\n",
        "TensorFlow also has built-in functions for MSE and MAE.\n",
        "\n",
        "For many evaluation functions, the premise is the same: compare predictions to the ground truth labels."
      ],
      "metadata": {
        "id": "04WWBaVJr1z2"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "# Calculate the mean absolute error\n",
        "mae = tf.keras.metrics.mean_absolute_error(y_test, y_preds)\n",
        "mae"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "QlVyxKJZsLGk",
        "outputId": "60d6a897-8a2d-427b-879f-f5a080a4dd17"
      },
      "execution_count": 32,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "<tf.Tensor: shape=(10,), dtype=float32, numpy=\n",
              "array([38.522648, 35.607407, 32.692165, 29.776922, 26.861689, 23.946445,\n",
              "       21.031204, 18.115967, 15.760577, 13.771291], dtype=float32)>"
            ]
          },
          "metadata": {},
          "execution_count": 32
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "It's showing in this format because the shape of `y_test` and `y_preds` are different, and to compare tensors, or do some operations, we have to make sure that shapes should be same."
      ],
      "metadata": {
        "id": "iFlmBrOysW-I"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "y_preds.shape, y_test.shape"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "NpaTxOzJsolN",
        "outputId": "43305a01-49d6-4c66-917e-c166e40accac"
      },
      "execution_count": 33,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "((10, 1), TensorShape([10]))"
            ]
          },
          "metadata": {},
          "execution_count": 33
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "# let's squeeze and then calculate mae\n",
        "mae = tf.keras.metrics.mean_absolute_error(y_test, tf.squeeze(y_preds))\n",
        "mae"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "V6RBNGfpsrRM",
        "outputId": "6ff2af9f-7898-4320-f468-8aa8b9bf0633"
      },
      "execution_count": 34,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "<tf.Tensor: shape=(), dtype=float32, numpy=25.404064>"
            ]
          },
          "metadata": {},
          "execution_count": 34
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "# calculate the mean squared error\n",
        "mse = tf.keras.metrics.mean_squared_error(y_test, tf.squeeze(y_preds))\n",
        "mse"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "OBCKUzuls295",
        "outputId": "382be139-b7ad-419d-eb24-1def45390c06"
      },
      "execution_count": 35,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "<tf.Tensor: shape=(), dtype=float32, numpy=655.07434>"
            ]
          },
          "metadata": {},
          "execution_count": 35
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "Create a function for each for later use"
      ],
      "metadata": {
        "id": "xxZuPQl-tAz8"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "def mae(y_test=y_test,y_preds=y_preds):\n",
        "  \"\"\"\n",
        "  Calculates mean absolute error between y_test and y_preds\n",
        "  \"\"\"\n",
        "  return tf.keras.metrics.mean_absolute_error(y_test, tf.squeeze(y_preds)).numpy()\n",
        "\n",
        "def mse(y_test=y_test,y_preds=y_preds):\n",
        "  \"\"\"\n",
        "  Calculates mean squared error between y_test and y_preds\n",
        "  \"\"\"\n",
        "  return tf.keras.metrics.mean_squared_error(y_test, tf.squeeze(y_preds)).numpy()"
      ],
      "metadata": {
        "id": "-Lv7EO38tGf2"
      },
      "execution_count": 36,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "source": [
        "We can also calculate Mean using pure TensorFlow function"
      ],
      "metadata": {
        "id": "Sa8TCOj5tkH9"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "tf.reduce_mean(tf.abs(y_test - y_preds.squeeze()))"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "v4AKB9n3twJP",
        "outputId": "4d3e2a59-3766-4906-d35a-d24ad02c5f5c"
      },
      "execution_count": 37,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "<tf.Tensor: shape=(), dtype=int32, numpy=25>"
            ]
          },
          "metadata": {},
          "execution_count": 37
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "## Running Experiments to Improve a model\n",
        "\n",
        "After seeing the evaluation metrics and the predictions your model makes, it's likely you'll want to improve it.\n",
        "\n",
        "Again, there are many different ways you can do this, but 3 of the main ones are:\n",
        "\n",
        "1. **Get more data** : get more examples for your model to train on (more opportunities to learn patterns).\n",
        "2. **Make your model larger (use a more complex model)** : this might come in the form of more layers or more hidden units in each layer.\n",
        "3. **Train for longer**: give your model more of a chance to find the patterns in the data\n",
        "\n",
        "Since we created our dataset, we could easily make more data but this isn't always the case when you're working with real-world datasets.\n",
        "\n",
        "So let's take a look at how we can improve our model using 2 and 3:\n",
        "\n",
        "To do so, we'll build 3 models and compare their results:\n",
        "\n",
        "1. `model_1`: same as original model, 1 layer, trained with 100 epochs.\n",
        "2. `model_2`: 2 layer, trained with 100 epochs\n",
        "3. `model_3`: 2 layer, trained with 500 epochs\n",
        "\n",
        "\n",
        "**Build `model_1`**"
      ],
      "metadata": {
        "id": "de4_MMTdt5y6"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "# set random seed\n",
        "tf.random.set_seed(4)\n",
        "\n",
        "# create a model\n",
        "model_1 = tf.keras.Sequential([\n",
        "    tf.keras.layers.Dense(1)\n",
        "])\n",
        "\n",
        "# compile a model\n",
        "model_1.compile(loss=tf.keras.losses.mae,\n",
        "                optimizer=tf.keras.optimizers.SGD(),\n",
        "                metrics=[\"mae\"])\n",
        "\n",
        "# fit the model\n",
        "model_1.fit(tf.expand_dims(X_train, axis=-1), y_train, epochs=100)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "QH-IFwM-0Xig",
        "outputId": "14d07128-8d74-4970-b0d2-6c8cc274d7e7"
      },
      "execution_count": 38,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Epoch 1/100\n",
            "2/2 [==============================] - 0s 6ms/step - loss: 26.2736 - mae: 26.2736\n",
            "Epoch 2/100\n",
            "2/2 [==============================] - 0s 7ms/step - loss: 5.8289 - mae: 5.8289\n",
            "Epoch 3/100\n",
            "2/2 [==============================] - 0s 9ms/step - loss: 5.3501 - mae: 5.3501\n",
            "Epoch 4/100\n",
            "2/2 [==============================] - 0s 4ms/step - loss: 11.0469 - mae: 11.0469\n",
            "Epoch 5/100\n",
            "2/2 [==============================] - 0s 9ms/step - loss: 6.0472 - mae: 6.0472\n",
            "Epoch 6/100\n",
            "2/2 [==============================] - 0s 4ms/step - loss: 4.8599 - mae: 4.8599\n",
            "Epoch 7/100\n",
            "2/2 [==============================] - 0s 5ms/step - loss: 8.4550 - mae: 8.4550\n",
            "Epoch 8/100\n",
            "2/2 [==============================] - 0s 5ms/step - loss: 9.1616 - mae: 9.1616\n",
            "Epoch 9/100\n",
            "2/2 [==============================] - 0s 4ms/step - loss: 5.4323 - mae: 5.4323\n",
            "Epoch 10/100\n",
            "2/2 [==============================] - 0s 5ms/step - loss: 8.8058 - mae: 8.8058\n",
            "Epoch 11/100\n",
            "2/2 [==============================] - 0s 8ms/step - loss: 11.9951 - mae: 11.9951\n",
            "Epoch 12/100\n",
            "2/2 [==============================] - 0s 6ms/step - loss: 9.9585 - mae: 9.9585\n",
            "Epoch 13/100\n",
            "2/2 [==============================] - 0s 9ms/step - loss: 12.5150 - mae: 12.5150\n",
            "Epoch 14/100\n",
            "2/2 [==============================] - 0s 4ms/step - loss: 5.2024 - mae: 5.2024\n",
            "Epoch 15/100\n",
            "2/2 [==============================] - 0s 4ms/step - loss: 12.8693 - mae: 12.8693\n",
            "Epoch 16/100\n",
            "2/2 [==============================] - 0s 5ms/step - loss: 8.1372 - mae: 8.1372\n",
            "Epoch 17/100\n",
            "2/2 [==============================] - 0s 8ms/step - loss: 4.8182 - mae: 4.8182\n",
            "Epoch 18/100\n",
            "2/2 [==============================] - 0s 12ms/step - loss: 8.9585 - mae: 8.9585\n",
            "Epoch 19/100\n",
            "2/2 [==============================] - 0s 6ms/step - loss: 5.3259 - mae: 5.3259\n",
            "Epoch 20/100\n",
            "2/2 [==============================] - 0s 4ms/step - loss: 6.8067 - mae: 6.8067\n",
            "Epoch 21/100\n",
            "2/2 [==============================] - 0s 5ms/step - loss: 7.2946 - mae: 7.2946\n",
            "Epoch 22/100\n",
            "2/2 [==============================] - 0s 5ms/step - loss: 7.9777 - mae: 7.9777\n",
            "Epoch 23/100\n",
            "2/2 [==============================] - 0s 5ms/step - loss: 7.8188 - mae: 7.8188\n",
            "Epoch 24/100\n",
            "2/2 [==============================] - 0s 5ms/step - loss: 10.2744 - mae: 10.2744\n",
            "Epoch 25/100\n",
            "2/2 [==============================] - 0s 5ms/step - loss: 12.4774 - mae: 12.4774\n",
            "Epoch 26/100\n",
            "2/2 [==============================] - 0s 9ms/step - loss: 5.9552 - mae: 5.9552\n",
            "Epoch 27/100\n",
            "2/2 [==============================] - 0s 5ms/step - loss: 5.6419 - mae: 5.6419\n",
            "Epoch 28/100\n",
            "2/2 [==============================] - 0s 4ms/step - loss: 11.6073 - mae: 11.6073\n",
            "Epoch 29/100\n",
            "2/2 [==============================] - 0s 4ms/step - loss: 12.1142 - mae: 12.1142\n",
            "Epoch 30/100\n",
            "2/2 [==============================] - 0s 4ms/step - loss: 9.3279 - mae: 9.3279\n",
            "Epoch 31/100\n",
            "2/2 [==============================] - 0s 4ms/step - loss: 10.8030 - mae: 10.8030\n",
            "Epoch 32/100\n",
            "2/2 [==============================] - 0s 5ms/step - loss: 16.1114 - mae: 16.1114\n",
            "Epoch 33/100\n",
            "2/2 [==============================] - 0s 4ms/step - loss: 9.3508 - mae: 9.3508\n",
            "Epoch 34/100\n",
            "2/2 [==============================] - 0s 5ms/step - loss: 6.9281 - mae: 6.9281\n",
            "Epoch 35/100\n",
            "2/2 [==============================] - 0s 6ms/step - loss: 4.7452 - mae: 4.7452\n",
            "Epoch 36/100\n",
            "2/2 [==============================] - 0s 5ms/step - loss: 5.0834 - mae: 5.0834\n",
            "Epoch 37/100\n",
            "2/2 [==============================] - 0s 7ms/step - loss: 5.4311 - mae: 5.4311\n",
            "Epoch 38/100\n",
            "2/2 [==============================] - 0s 6ms/step - loss: 5.4156 - mae: 5.4156\n",
            "Epoch 39/100\n",
            "2/2 [==============================] - 0s 6ms/step - loss: 5.2098 - mae: 5.2098\n",
            "Epoch 40/100\n",
            "2/2 [==============================] - 0s 7ms/step - loss: 12.3448 - mae: 12.3448\n",
            "Epoch 41/100\n",
            "2/2 [==============================] - 0s 6ms/step - loss: 8.4752 - mae: 8.4752\n",
            "Epoch 42/100\n",
            "2/2 [==============================] - 0s 6ms/step - loss: 11.0883 - mae: 11.0883\n",
            "Epoch 43/100\n",
            "2/2 [==============================] - 0s 7ms/step - loss: 9.7070 - mae: 9.7070\n",
            "Epoch 44/100\n",
            "2/2 [==============================] - 0s 5ms/step - loss: 15.1892 - mae: 15.1892\n",
            "Epoch 45/100\n",
            "2/2 [==============================] - 0s 6ms/step - loss: 6.9533 - mae: 6.9533\n",
            "Epoch 46/100\n",
            "2/2 [==============================] - 0s 5ms/step - loss: 5.8549 - mae: 5.8549\n",
            "Epoch 47/100\n",
            "2/2 [==============================] - 0s 6ms/step - loss: 13.6988 - mae: 13.6988\n",
            "Epoch 48/100\n",
            "2/2 [==============================] - 0s 6ms/step - loss: 10.2165 - mae: 10.2165\n",
            "Epoch 49/100\n",
            "2/2 [==============================] - 0s 5ms/step - loss: 11.3079 - mae: 11.3079\n",
            "Epoch 50/100\n",
            "2/2 [==============================] - 0s 5ms/step - loss: 8.1985 - mae: 8.1985\n",
            "Epoch 51/100\n",
            "2/2 [==============================] - 0s 5ms/step - loss: 10.7097 - mae: 10.7097\n",
            "Epoch 52/100\n",
            "2/2 [==============================] - 0s 5ms/step - loss: 5.8176 - mae: 5.8176\n",
            "Epoch 53/100\n",
            "2/2 [==============================] - 0s 5ms/step - loss: 7.1681 - mae: 7.1681\n",
            "Epoch 54/100\n",
            "2/2 [==============================] - 0s 5ms/step - loss: 7.7591 - mae: 7.7591\n",
            "Epoch 55/100\n",
            "2/2 [==============================] - 0s 5ms/step - loss: 8.7092 - mae: 8.7092\n",
            "Epoch 56/100\n",
            "2/2 [==============================] - 0s 9ms/step - loss: 7.1068 - mae: 7.1068\n",
            "Epoch 57/100\n",
            "2/2 [==============================] - 0s 10ms/step - loss: 11.7226 - mae: 11.7226\n",
            "Epoch 58/100\n",
            "2/2 [==============================] - 0s 11ms/step - loss: 5.6130 - mae: 5.6130\n",
            "Epoch 59/100\n",
            "2/2 [==============================] - 0s 9ms/step - loss: 7.1943 - mae: 7.1943\n",
            "Epoch 60/100\n",
            "2/2 [==============================] - 0s 8ms/step - loss: 12.7965 - mae: 12.7965\n",
            "Epoch 61/100\n",
            "2/2 [==============================] - 0s 8ms/step - loss: 8.1668 - mae: 8.1668\n",
            "Epoch 62/100\n",
            "2/2 [==============================] - 0s 6ms/step - loss: 7.5795 - mae: 7.5795\n",
            "Epoch 63/100\n",
            "2/2 [==============================] - 0s 11ms/step - loss: 10.9615 - mae: 10.9615\n",
            "Epoch 64/100\n",
            "2/2 [==============================] - 0s 9ms/step - loss: 12.3831 - mae: 12.3831\n",
            "Epoch 65/100\n",
            "2/2 [==============================] - 0s 4ms/step - loss: 6.4478 - mae: 6.4478\n",
            "Epoch 66/100\n",
            "2/2 [==============================] - 0s 4ms/step - loss: 4.9967 - mae: 4.9967\n",
            "Epoch 67/100\n",
            "2/2 [==============================] - 0s 7ms/step - loss: 6.1890 - mae: 6.1890\n",
            "Epoch 68/100\n",
            "2/2 [==============================] - 0s 5ms/step - loss: 10.1926 - mae: 10.1926\n",
            "Epoch 69/100\n",
            "2/2 [==============================] - 0s 5ms/step - loss: 8.7438 - mae: 8.7438\n",
            "Epoch 70/100\n",
            "2/2 [==============================] - 0s 4ms/step - loss: 5.7691 - mae: 5.7691\n",
            "Epoch 71/100\n",
            "2/2 [==============================] - 0s 8ms/step - loss: 8.6417 - mae: 8.6417\n",
            "Epoch 72/100\n",
            "2/2 [==============================] - 0s 8ms/step - loss: 19.4683 - mae: 19.4683\n",
            "Epoch 73/100\n",
            "2/2 [==============================] - 0s 5ms/step - loss: 6.5358 - mae: 6.5358\n",
            "Epoch 74/100\n",
            "2/2 [==============================] - 0s 12ms/step - loss: 3.4672 - mae: 3.4672\n",
            "Epoch 75/100\n",
            "2/2 [==============================] - 0s 7ms/step - loss: 13.5750 - mae: 13.5750\n",
            "Epoch 76/100\n",
            "2/2 [==============================] - 0s 4ms/step - loss: 9.1122 - mae: 9.1122\n",
            "Epoch 77/100\n",
            "2/2 [==============================] - 0s 6ms/step - loss: 6.2883 - mae: 6.2883\n",
            "Epoch 78/100\n",
            "2/2 [==============================] - 0s 6ms/step - loss: 6.9682 - mae: 6.9682\n",
            "Epoch 79/100\n",
            "2/2 [==============================] - 0s 4ms/step - loss: 5.2651 - mae: 5.2651\n",
            "Epoch 80/100\n",
            "2/2 [==============================] - 0s 5ms/step - loss: 4.4525 - mae: 4.4525\n",
            "Epoch 81/100\n",
            "2/2 [==============================] - 0s 4ms/step - loss: 4.4769 - mae: 4.4769\n",
            "Epoch 82/100\n",
            "2/2 [==============================] - 0s 5ms/step - loss: 11.3614 - mae: 11.3614\n",
            "Epoch 83/100\n",
            "2/2 [==============================] - 0s 5ms/step - loss: 11.4737 - mae: 11.4737\n",
            "Epoch 84/100\n",
            "2/2 [==============================] - 0s 7ms/step - loss: 5.7549 - mae: 5.7549\n",
            "Epoch 85/100\n",
            "2/2 [==============================] - 0s 5ms/step - loss: 9.7125 - mae: 9.7125\n",
            "Epoch 86/100\n",
            "2/2 [==============================] - 0s 7ms/step - loss: 14.0676 - mae: 14.0676\n",
            "Epoch 87/100\n",
            "2/2 [==============================] - 0s 4ms/step - loss: 6.5628 - mae: 6.5628\n",
            "Epoch 88/100\n",
            "2/2 [==============================] - 0s 5ms/step - loss: 8.5756 - mae: 8.5756\n",
            "Epoch 89/100\n",
            "2/2 [==============================] - 0s 5ms/step - loss: 13.3504 - mae: 13.3504\n",
            "Epoch 90/100\n",
            "2/2 [==============================] - 0s 4ms/step - loss: 8.5381 - mae: 8.5381\n",
            "Epoch 91/100\n",
            "2/2 [==============================] - 0s 6ms/step - loss: 15.7728 - mae: 15.7728\n",
            "Epoch 92/100\n",
            "2/2 [==============================] - 0s 10ms/step - loss: 12.5162 - mae: 12.5162\n",
            "Epoch 93/100\n",
            "2/2 [==============================] - 0s 6ms/step - loss: 6.0417 - mae: 6.0417\n",
            "Epoch 94/100\n",
            "2/2 [==============================] - 0s 5ms/step - loss: 8.3045 - mae: 8.3045\n",
            "Epoch 95/100\n",
            "2/2 [==============================] - 0s 7ms/step - loss: 10.9778 - mae: 10.9778\n",
            "Epoch 96/100\n",
            "2/2 [==============================] - 0s 6ms/step - loss: 5.6015 - mae: 5.6015\n",
            "Epoch 97/100\n",
            "2/2 [==============================] - 0s 6ms/step - loss: 16.0907 - mae: 16.0907\n",
            "Epoch 98/100\n",
            "2/2 [==============================] - 0s 9ms/step - loss: 6.8727 - mae: 6.8727\n",
            "Epoch 99/100\n",
            "2/2 [==============================] - 0s 7ms/step - loss: 6.2880 - mae: 6.2880\n",
            "Epoch 100/100\n",
            "2/2 [==============================] - 0s 5ms/step - loss: 10.3345 - mae: 10.3345\n"
          ]
        },
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "<keras.callbacks.History at 0x7fe5b7711250>"
            ]
          },
          "metadata": {},
          "execution_count": 38
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "# make predictions\n",
        "y_pred_1 = model_1.predict(X_test)\n",
        "y_pred_1"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "5Y9bm-fr06nc",
        "outputId": "361ce96d-738d-4f70-a234-7742f5447a80"
      },
      "execution_count": 39,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "WARNING:tensorflow:5 out of the last 5 calls to <function Model.make_predict_function.<locals>.predict_function at 0x7fe5bd1530e0> triggered tf.function retracing. Tracing is expensive and the excessive number of tracings could be due to (1) creating @tf.function repeatedly in a loop, (2) passing tensors with different shapes, (3) passing Python objects instead of tensors. For (1), please define your @tf.function outside of the loop. For (2), @tf.function has reduce_retracing=True option that can avoid unnecessary retracing. For (3), please refer to https://www.tensorflow.org/guide/function#controlling_retracing and https://www.tensorflow.org/api_docs/python/tf/function for  more details.\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "1/1 [==============================] - 0s 44ms/step\n"
          ]
        },
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "array([[50.149574],\n",
              "       [53.43563 ],\n",
              "       [56.721687],\n",
              "       [60.00774 ],\n",
              "       [63.293797],\n",
              "       [66.57985 ],\n",
              "       [69.865906],\n",
              "       [73.15196 ],\n",
              "       [76.43802 ],\n",
              "       [79.724075]], dtype=float32)"
            ]
          },
          "metadata": {},
          "execution_count": 39
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "# plot predictions\n",
        "plot_prediction(predictions=y_pred_1)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 428
        },
        "id": "LrenoraM1Wtg",
        "outputId": "5fd4e164-3b9b-4b3d-9d2e-0f11e1730e43"
      },
      "execution_count": 40,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 720x504 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "# check metrics\n",
        "mae_1 = mae(y_test,y_pred_1)\n",
        "mae_1"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "7rAh7vCf1eRJ",
        "outputId": "79d32961-a840-45b1-8cde-6f134fe76a03"
      },
      "execution_count": 41,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "18.063175"
            ]
          },
          "metadata": {},
          "execution_count": 41
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "mse_1 = mse(y_test,y_pred_1)\n",
        "mse_1"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "F1RuVE121rPQ",
        "outputId": "33a1c9db-8c53-4f7a-854d-0bfa37409085"
      },
      "execution_count": 42,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "330.4835"
            ]
          },
          "metadata": {},
          "execution_count": 42
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "**Build `model_2`**\n",
        "\n",
        "Adding one extra dense layer"
      ],
      "metadata": {
        "id": "sd8ni6UT13_5"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "# set random seed\n",
        "tf.random.set_seed(42)\n",
        "\n",
        "# create a model\n",
        "model_2 = tf.keras.Sequential([\n",
        "    tf.keras.layers.Dense(1),\n",
        "    tf.keras.layers.Dense(1) # add a second layer\n",
        "])\n",
        "\n",
        "# compile a model\n",
        "model_2.compile(loss=tf.keras.losses.mae,\n",
        "                optimizer=tf.keras.optimizers.SGD(),\n",
        "                metrics=[\"mae\"])\n",
        "\n",
        "# fit the model\n",
        "model_2.fit(tf.expand_dims(X_train, axis=-1), y_train, epochs = 100, verbose=0)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "go2l6ZeT2CL8",
        "outputId": "0e2a1302-477c-44c0-e5ae-654482cbe8c1"
      },
      "execution_count": 43,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "<keras.callbacks.History at 0x7fe5b85ed710>"
            ]
          },
          "metadata": {},
          "execution_count": 43
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "# Make predictions and plot model_2\n",
        "y_preds_2 = model_2.predict(X_test)\n",
        "plot_prediction(predictions=y_preds_2)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 486
        },
        "id": "3u1ktti74f1W",
        "outputId": "ba6a28db-2930-4d7c-f17f-e34b55ad853f"
      },
      "execution_count": 44,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "WARNING:tensorflow:6 out of the last 6 calls to <function Model.make_predict_function.<locals>.predict_function at 0x7fe5b8531320> triggered tf.function retracing. Tracing is expensive and the excessive number of tracings could be due to (1) creating @tf.function repeatedly in a loop, (2) passing tensors with different shapes, (3) passing Python objects instead of tensors. For (1), please define your @tf.function outside of the loop. For (2), @tf.function has reduce_retracing=True option that can avoid unnecessary retracing. For (3), please refer to https://www.tensorflow.org/guide/function#controlling_retracing and https://www.tensorflow.org/api_docs/python/tf/function for  more details.\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "1/1 [==============================] - 0s 54ms/step\n"
          ]
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 720x504 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "# Calculate model_2 metrics\n",
        "mae_2 = mae(y_preds=y_preds_2)\n",
        "mse_2 = mse(y_preds=y_preds_2)\n",
        "mae_2, mse_2"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "gzxJldE-4xjH",
        "outputId": "4bb2fa6c-cf70-42f1-c75b-db9ad83f6339"
      },
      "execution_count": 45,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "(30.45366, 953.075)"
            ]
          },
          "metadata": {},
          "execution_count": 45
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "**Build `model_3`**\n",
        "\n",
        "2 layer, trained with 500 epochs"
      ],
      "metadata": {
        "id": "WhUH2l3x5Js5"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "# set the random seed\n",
        "tf.random.set_seed(42)\n",
        "\n",
        "#1. Create a model\n",
        "model_3 = tf.keras.Sequential([\n",
        "    tf.keras.layers.Dense(10),\n",
        "    tf.keras.layers.Dense(1)\n",
        "])\n",
        "\n",
        "# 2. Compile a model\n",
        "model_3.compile(loss=tf.keras.losses.mae,\n",
        "                optimizer=tf.keras.optimizers.SGD(),\n",
        "                metrics=[\"mae\"])\n",
        "\n",
        "# 3. Fit a model\n",
        "model_3.fit(tf.expand_dims(X_train, axis=-1), y_train, epochs=500, verbose = 0)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "SyPTfS8xqGJn",
        "outputId": "ab31a909-95b6-43dd-c5c8-e23ce5ec1965"
      },
      "execution_count": 47,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "<keras.callbacks.History at 0x7fe5b66f29d0>"
            ]
          },
          "metadata": {},
          "execution_count": 47
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "# Make predictions and plot model_3\n",
        "y_preds_3 = model_3.predict(X_test)\n",
        "plot_prediction(predictions=y_preds_3)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 445
        },
        "id": "hx3ePH5UqwLt",
        "outputId": "0d83f960-2910-454c-9162-73c6684899ee"
      },
      "execution_count": 48,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "1/1 [==============================] - 0s 71ms/step\n"
          ]
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 720x504 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "Even worse 😆"
      ],
      "metadata": {
        "id": "3aAiQqzJrEGa"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "# lets check mae and mse\n",
        "mae_3 = mae(y_test, y_preds_3)\n",
        "mse_3 = mse(y_test, y_preds_3)\n",
        "mae_3, mse_3"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "Y3oUYIiHrJTT",
        "outputId": "c9a374a4-aebc-45b1-c468-4a36d2291a35"
      },
      "execution_count": 49,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "(66.00897, 4443.747)"
            ]
          },
          "metadata": {},
          "execution_count": 49
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "So how can we know which experiment went well, we have visualized it, but let's compare each metrics of each model"
      ],
      "metadata": {
        "id": "ZS2vKjpJrXXr"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "# creating a dataframe\n",
        "\n",
        "import pandas as pd\n",
        "d = [[mae_1, mae_2, mae_3], [mse_1, mse_2, mse_3]]\n",
        "metrics_df = pd.DataFrame(data=d ,index=[\"MAE\", \"MSE\"],columns=[\"Model 1\",\"Model 2\", \"Model 3\"])\n",
        "metrics_df"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 112
        },
        "id": "-DRl_ocark1K",
        "outputId": "c8dc3dda-c9ed-431a-e893-02dd06dcf4f0"
      },
      "execution_count": 52,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "        Model 1     Model 2      Model 3\n",
              "MAE   18.063175   30.453659    66.008972\n",
              "MSE  330.483490  953.075012  4443.747070"
            ],
            "text/html": [
              "\n",
              "  <div id=\"df-87cdb250-7b20-490d-9973-1a28ca996a40\">\n",
              "    <div class=\"colab-df-container\">\n",
              "      <div>\n",
              "<style scoped>\n",
              "    .dataframe tbody tr th:only-of-type {\n",
              "        vertical-align: middle;\n",
              "    }\n",
              "\n",
              "    .dataframe tbody tr th {\n",
              "        vertical-align: top;\n",
              "    }\n",
              "\n",
              "    .dataframe thead th {\n",
              "        text-align: right;\n",
              "    }\n",
              "</style>\n",
              "<table border=\"1\" class=\"dataframe\">\n",
              "  <thead>\n",
              "    <tr style=\"text-align: right;\">\n",
              "      <th></th>\n",
              "      <th>Model 1</th>\n",
              "      <th>Model 2</th>\n",
              "      <th>Model 3</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>MAE</th>\n",
              "      <td>18.063175</td>\n",
              "      <td>30.453659</td>\n",
              "      <td>66.008972</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>MSE</th>\n",
              "      <td>330.483490</td>\n",
              "      <td>953.075012</td>\n",
              "      <td>4443.747070</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>\n",
              "      <button class=\"colab-df-convert\" onclick=\"convertToInteractive('df-87cdb250-7b20-490d-9973-1a28ca996a40')\"\n",
              "              title=\"Convert this dataframe to an interactive table.\"\n",
              "              style=\"display:none;\">\n",
              "        \n",
              "  <svg xmlns=\"http://www.w3.org/2000/svg\" height=\"24px\"viewBox=\"0 0 24 24\"\n",
              "       width=\"24px\">\n",
              "    <path d=\"M0 0h24v24H0V0z\" fill=\"none\"/>\n",
              "    <path d=\"M18.56 5.44l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94zm-11 1L8.5 8.5l.94-2.06 2.06-.94-2.06-.94L8.5 2.5l-.94 2.06-2.06.94zm10 10l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94z\"/><path d=\"M17.41 7.96l-1.37-1.37c-.4-.4-.92-.59-1.43-.59-.52 0-1.04.2-1.43.59L10.3 9.45l-7.72 7.72c-.78.78-.78 2.05 0 2.83L4 21.41c.39.39.9.59 1.41.59.51 0 1.02-.2 1.41-.59l7.78-7.78 2.81-2.81c.8-.78.8-2.07 0-2.86zM5.41 20L4 18.59l7.72-7.72 1.47 1.35L5.41 20z\"/>\n",
              "  </svg>\n",
              "      </button>\n",
              "      \n",
              "  <style>\n",
              "    .colab-df-container {\n",
              "      display:flex;\n",
              "      flex-wrap:wrap;\n",
              "      gap: 12px;\n",
              "    }\n",
              "\n",
              "    .colab-df-convert {\n",
              "      background-color: #E8F0FE;\n",
              "      border: none;\n",
              "      border-radius: 50%;\n",
              "      cursor: pointer;\n",
              "      display: none;\n",
              "      fill: #1967D2;\n",
              "      height: 32px;\n",
              "      padding: 0 0 0 0;\n",
              "      width: 32px;\n",
              "    }\n",
              "\n",
              "    .colab-df-convert:hover {\n",
              "      background-color: #E2EBFA;\n",
              "      box-shadow: 0px 1px 2px rgba(60, 64, 67, 0.3), 0px 1px 3px 1px rgba(60, 64, 67, 0.15);\n",
              "      fill: #174EA6;\n",
              "    }\n",
              "\n",
              "    [theme=dark] .colab-df-convert {\n",
              "      background-color: #3B4455;\n",
              "      fill: #D2E3FC;\n",
              "    }\n",
              "\n",
              "    [theme=dark] .colab-df-convert:hover {\n",
              "      background-color: #434B5C;\n",
              "      box-shadow: 0px 1px 3px 1px rgba(0, 0, 0, 0.15);\n",
              "      filter: drop-shadow(0px 1px 2px rgba(0, 0, 0, 0.3));\n",
              "      fill: #FFFFFF;\n",
              "    }\n",
              "  </style>\n",
              "\n",
              "      <script>\n",
              "        const buttonEl =\n",
              "          document.querySelector('#df-87cdb250-7b20-490d-9973-1a28ca996a40 button.colab-df-convert');\n",
              "        buttonEl.style.display =\n",
              "          google.colab.kernel.accessAllowed ? 'block' : 'none';\n",
              "\n",
              "        async function convertToInteractive(key) {\n",
              "          const element = document.querySelector('#df-87cdb250-7b20-490d-9973-1a28ca996a40');\n",
              "          const dataTable =\n",
              "            await google.colab.kernel.invokeFunction('convertToInteractive',\n",
              "                                                     [key], {});\n",
              "          if (!dataTable) return;\n",
              "\n",
              "          const docLinkHtml = 'Like what you see? Visit the ' +\n",
              "            '<a target=\"_blank\" href=https://colab.research.google.com/notebooks/data_table.ipynb>data table notebook</a>'\n",
              "            + ' to learn more about interactive tables.';\n",
              "          element.innerHTML = '';\n",
              "          dataTable['output_type'] = 'display_data';\n",
              "          await google.colab.output.renderOutput(dataTable, element);\n",
              "          const docLink = document.createElement('div');\n",
              "          docLink.innerHTML = docLinkHtml;\n",
              "          element.appendChild(docLink);\n",
              "        }\n",
              "      </script>\n",
              "    </div>\n",
              "  </div>\n",
              "  "
            ]
          },
          "metadata": {},
          "execution_count": 52
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "# if we plot a bar graph\n",
        "metrics_df.plot(kind=\"bar\")"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 294
        },
        "id": "uqjcgFNgrwbT",
        "outputId": "b3c04d53-1083-478a-8321-a4700d6e5954"
      },
      "execution_count": 56,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "<matplotlib.axes._subplots.AxesSubplot at 0x7fe5b118cf10>"
            ]
          },
          "metadata": {},
          "execution_count": 56
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "So from our experiments it looks like `model_1` performed the best.\n",
        "\n",
        "But usually this is very less number of experiments and changes we did to adjust the hyperparameters and increase layers of neural network for experiments.\n",
        "\n",
        "> 🔑 **Note**: One of your main goals should be to minimize the time between your experiments. The more experiments you do, the more things you'll figure out which don't work and in turn, get closer to figuring out what does work. Remember the machine learning practitioner's motto: **\"Experiment\", \"Experiment\" and \"Experiment\"**."
      ],
      "metadata": {
        "id": "iLWzA1V_thXU"
      }
    },
    {
      "cell_type": "markdown",
      "source": [
        "## Tracking our experiments\n",
        "\n",
        "To track our experiments we have used variables and stored the results into them and visualized it, but there are more tools and libraries available for this.\n",
        "\n",
        "> 📖 **Resource**:\n",
        "  1. [TensorBoard](https://tensorboard.dev/) - a component of TensorFlow library to help track modelling experiments.\n",
        "  2. [Weights & Biases](https://www.wandb.com/) - a tool for tracking all kinds of machine learning experiments."
      ],
      "metadata": {
        "id": "X7w6imq0u21B"
      }
    },
    {
      "cell_type": "markdown",
      "source": [
        "## Saving a model\n",
        "\n",
        "Once we've trained a model and found one which performs to your liking, we'll probabily want to save it for use elsewhere(like a web application or mobile devices).\n",
        "\n",
        "We can save a TensorFlow/Keras model using `model.save()`\n",
        "\n",
        "There are two ways to save a model in TensorFlow\n",
        "1. `SavedModel` format(default)\n",
        "2. `HDF5` format."
      ],
      "metadata": {
        "id": "qU7g_ej7vfYu"
      }
    },
    {
      "cell_type": "markdown",
      "source": [
        "let's save it using `SavedModel`"
      ],
      "metadata": {
        "id": "PGAxrHizxA2U"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "model_1.save(\"best_model_SavedModel_format_regression_NN\")"
      ],
      "metadata": {
        "id": "7-O5u0ZexFR3"
      },
      "execution_count": 57,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "# check it out\n",
        "!ls best_model_SavedModel_format_regression_NN"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "18EWFHLgxPIB",
        "outputId": "0ef6344d-9569-4c2f-e261-c9b3f95f0ab0"
      },
      "execution_count": 58,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "assets\tkeras_metadata.pb  saved_model.pb  variables\n"
          ]
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "using `HDF5` format"
      ],
      "metadata": {
        "id": "7IzIpg5jxXf8"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "model_1.save(\"best_model_HDF5_format.h5\")"
      ],
      "metadata": {
        "id": "n_ssHmEWxewK"
      },
      "execution_count": 59,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "!ls best_model_HDF5_format.h5"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "bgH2xNr1xkX0",
        "outputId": "cd1230f5-a4f4-4a19-bc26-0378189256ab"
      },
      "execution_count": 60,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "best_model_HDF5_format.h5\n"
          ]
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "## Loading a model\n",
        "\n",
        "We can load a saved model using the `load_model()` method"
      ],
      "metadata": {
        "id": "k4Mw1qpuxoJ9"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "model_loaded = tf.keras.models.load_model(\"best_model_HDF5_format.h5\")\n",
        "model_loaded.summary()"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "JTcush_wxy-7",
        "outputId": "6ddaf997-b0d4-4c0c-a521-1710aa131b77"
      },
      "execution_count": 61,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Model: \"sequential_5\"\n",
            "_________________________________________________________________\n",
            " Layer (type)                Output Shape              Param #   \n",
            "=================================================================\n",
            " dense_5 (Dense)             (None, 1)                 2         \n",
            "                                                                 \n",
            "=================================================================\n",
            "Total params: 2\n",
            "Trainable params: 2\n",
            "Non-trainable params: 0\n",
            "_________________________________________________________________\n"
          ]
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "## Downloading a model (from Google colab)\n",
        "\n",
        "We can download it manually using right-click over the file and then Download it.\n",
        "\n",
        "But using code also it's possible."
      ],
      "metadata": {
        "id": "h8yeqrqzx82D"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "from google.colab import files\n",
        "files.download(\"best_model_SavedModel_format_regression_NN\")"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 34
        },
        "id": "AlQvLMCWyQWW",
        "outputId": "8f22e4ff-23d9-4833-c778-836717186901"
      },
      "execution_count": 62,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<IPython.core.display.Javascript object>"
            ],
            "application/javascript": [
              "\n",
              "    async function download(id, filename, size) {\n",
              "      if (!google.colab.kernel.accessAllowed) {\n",
              "        return;\n",
              "      }\n",
              "      const div = document.createElement('div');\n",
              "      const label = document.createElement('label');\n",
              "      label.textContent = `Downloading \"${filename}\": `;\n",
              "      div.appendChild(label);\n",
              "      const progress = document.createElement('progress');\n",
              "      progress.max = size;\n",
              "      div.appendChild(progress);\n",
              "      document.body.appendChild(div);\n",
              "\n",
              "      const buffers = [];\n",
              "      let downloaded = 0;\n",
              "\n",
              "      const channel = await google.colab.kernel.comms.open(id);\n",
              "      // Send a message to notify the kernel that we're ready.\n",
              "      channel.send({})\n",
              "\n",
              "      for await (const message of channel.messages) {\n",
              "        // Send a message to notify the kernel that we're ready.\n",
              "        channel.send({})\n",
              "        if (message.buffers) {\n",
              "          for (const buffer of message.buffers) {\n",
              "            buffers.push(buffer);\n",
              "            downloaded += buffer.byteLength;\n",
              "            progress.value = downloaded;\n",
              "          }\n",
              "        }\n",
              "      }\n",
              "      const blob = new Blob(buffers, {type: 'application/binary'});\n",
              "      const a = document.createElement('a');\n",
              "      a.href = window.URL.createObjectURL(blob);\n",
              "      a.download = filename;\n",
              "      div.appendChild(a);\n",
              "      a.click();\n",
              "      div.remove();\n",
              "    }\n",
              "  "
            ]
          },
          "metadata": {}
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<IPython.core.display.Javascript object>"
            ],
            "application/javascript": [
              "download(\"download_ad67a8f7-b09c-4b2c-a739-b7458dbb5b61\", \"best_model_SavedModel_format_regression_NN\", 4096)"
            ]
          },
          "metadata": {}
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "## A Larger Example\n",
        "\n",
        "We've seen all the fundamentals of building neural network regression models in TensorFlow\n",
        "\n",
        "Let's step it up a notch and build a model for a more feature rich dataset.\n",
        "\n",
        "\n",
        "For this we will use a dataset available in Kaggle i.e [Medical Cost dataset](https://www.kaggle.com/mirichoi0218/insurance), [hosted on Github](https://github.com/stedy/Machine-Learning-with-R-datasets/blob/master/insurance.csv)"
      ],
      "metadata": {
        "id": "o-KLHzXgyZye"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "# Import required libraries\n",
        "import pandas as pd\n",
        "import matplotlib.pyplot as plt\n",
        "import tensorflow as tf"
      ],
      "metadata": {
        "id": "MdYNf0IAzBZ0"
      },
      "execution_count": 63,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "source": [
        "The problem we're solving through dataset is:\n",
        "We'll try to predict the cost of the health insurance based on given data like `age`, `sex`, `smoker`, `region`, `bmi`, `children`"
      ],
      "metadata": {
        "id": "jSoVbNGDzP3q"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "# import the dataset\n",
        "insurance = pd.read_csv(\"https://raw.githubusercontent.com/stedy/Machine-Learning-with-R-datasets/master/insurance.csv\")"
      ],
      "metadata": {
        "id": "QVmBwj8RzHoi"
      },
      "execution_count": 65,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "insurance.head()"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 206
        },
        "id": "GCsrjQoMzjwd",
        "outputId": "b6b17b21-133e-471f-b5be-f8761bae6b26"
      },
      "execution_count": 66,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "   age     sex     bmi  children smoker     region      charges\n",
              "0   19  female  27.900         0    yes  southwest  16884.92400\n",
              "1   18    male  33.770         1     no  southeast   1725.55230\n",
              "2   28    male  33.000         3     no  southeast   4449.46200\n",
              "3   33    male  22.705         0     no  northwest  21984.47061\n",
              "4   32    male  28.880         0     no  northwest   3866.85520"
            ],
            "text/html": [
              "\n",
              "  <div id=\"df-afa8d751-827f-4254-bcaf-c3d6081a9b45\">\n",
              "    <div class=\"colab-df-container\">\n",
              "      <div>\n",
              "<style scoped>\n",
              "    .dataframe tbody tr th:only-of-type {\n",
              "        vertical-align: middle;\n",
              "    }\n",
              "\n",
              "    .dataframe tbody tr th {\n",
              "        vertical-align: top;\n",
              "    }\n",
              "\n",
              "    .dataframe thead th {\n",
              "        text-align: right;\n",
              "    }\n",
              "</style>\n",
              "<table border=\"1\" class=\"dataframe\">\n",
              "  <thead>\n",
              "    <tr style=\"text-align: right;\">\n",
              "      <th></th>\n",
              "      <th>age</th>\n",
              "      <th>sex</th>\n",
              "      <th>bmi</th>\n",
              "      <th>children</th>\n",
              "      <th>smoker</th>\n",
              "      <th>region</th>\n",
              "      <th>charges</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>19</td>\n",
              "      <td>female</td>\n",
              "      <td>27.900</td>\n",
              "      <td>0</td>\n",
              "      <td>yes</td>\n",
              "      <td>southwest</td>\n",
              "      <td>16884.92400</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>18</td>\n",
              "      <td>male</td>\n",
              "      <td>33.770</td>\n",
              "      <td>1</td>\n",
              "      <td>no</td>\n",
              "      <td>southeast</td>\n",
              "      <td>1725.55230</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>28</td>\n",
              "      <td>male</td>\n",
              "      <td>33.000</td>\n",
              "      <td>3</td>\n",
              "      <td>no</td>\n",
              "      <td>southeast</td>\n",
              "      <td>4449.46200</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3</th>\n",
              "      <td>33</td>\n",
              "      <td>male</td>\n",
              "      <td>22.705</td>\n",
              "      <td>0</td>\n",
              "      <td>no</td>\n",
              "      <td>northwest</td>\n",
              "      <td>21984.47061</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4</th>\n",
              "      <td>32</td>\n",
              "      <td>male</td>\n",
              "      <td>28.880</td>\n",
              "      <td>0</td>\n",
              "      <td>no</td>\n",
              "      <td>northwest</td>\n",
              "      <td>3866.85520</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>\n",
              "      <button class=\"colab-df-convert\" onclick=\"convertToInteractive('df-afa8d751-827f-4254-bcaf-c3d6081a9b45')\"\n",
              "              title=\"Convert this dataframe to an interactive table.\"\n",
              "              style=\"display:none;\">\n",
              "        \n",
              "  <svg xmlns=\"http://www.w3.org/2000/svg\" height=\"24px\"viewBox=\"0 0 24 24\"\n",
              "       width=\"24px\">\n",
              "    <path d=\"M0 0h24v24H0V0z\" fill=\"none\"/>\n",
              "    <path d=\"M18.56 5.44l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94zm-11 1L8.5 8.5l.94-2.06 2.06-.94-2.06-.94L8.5 2.5l-.94 2.06-2.06.94zm10 10l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94z\"/><path d=\"M17.41 7.96l-1.37-1.37c-.4-.4-.92-.59-1.43-.59-.52 0-1.04.2-1.43.59L10.3 9.45l-7.72 7.72c-.78.78-.78 2.05 0 2.83L4 21.41c.39.39.9.59 1.41.59.51 0 1.02-.2 1.41-.59l7.78-7.78 2.81-2.81c.8-.78.8-2.07 0-2.86zM5.41 20L4 18.59l7.72-7.72 1.47 1.35L5.41 20z\"/>\n",
              "  </svg>\n",
              "      </button>\n",
              "      \n",
              "  <style>\n",
              "    .colab-df-container {\n",
              "      display:flex;\n",
              "      flex-wrap:wrap;\n",
              "      gap: 12px;\n",
              "    }\n",
              "\n",
              "    .colab-df-convert {\n",
              "      background-color: #E8F0FE;\n",
              "      border: none;\n",
              "      border-radius: 50%;\n",
              "      cursor: pointer;\n",
              "      display: none;\n",
              "      fill: #1967D2;\n",
              "      height: 32px;\n",
              "      padding: 0 0 0 0;\n",
              "      width: 32px;\n",
              "    }\n",
              "\n",
              "    .colab-df-convert:hover {\n",
              "      background-color: #E2EBFA;\n",
              "      box-shadow: 0px 1px 2px rgba(60, 64, 67, 0.3), 0px 1px 3px 1px rgba(60, 64, 67, 0.15);\n",
              "      fill: #174EA6;\n",
              "    }\n",
              "\n",
              "    [theme=dark] .colab-df-convert {\n",
              "      background-color: #3B4455;\n",
              "      fill: #D2E3FC;\n",
              "    }\n",
              "\n",
              "    [theme=dark] .colab-df-convert:hover {\n",
              "      background-color: #434B5C;\n",
              "      box-shadow: 0px 1px 3px 1px rgba(0, 0, 0, 0.15);\n",
              "      filter: drop-shadow(0px 1px 2px rgba(0, 0, 0, 0.3));\n",
              "      fill: #FFFFFF;\n",
              "    }\n",
              "  </style>\n",
              "\n",
              "      <script>\n",
              "        const buttonEl =\n",
              "          document.querySelector('#df-afa8d751-827f-4254-bcaf-c3d6081a9b45 button.colab-df-convert');\n",
              "        buttonEl.style.display =\n",
              "          google.colab.kernel.accessAllowed ? 'block' : 'none';\n",
              "\n",
              "        async function convertToInteractive(key) {\n",
              "          const element = document.querySelector('#df-afa8d751-827f-4254-bcaf-c3d6081a9b45');\n",
              "          const dataTable =\n",
              "            await google.colab.kernel.invokeFunction('convertToInteractive',\n",
              "                                                     [key], {});\n",
              "          if (!dataTable) return;\n",
              "\n",
              "          const docLinkHtml = 'Like what you see? Visit the ' +\n",
              "            '<a target=\"_blank\" href=https://colab.research.google.com/notebooks/data_table.ipynb>data table notebook</a>'\n",
              "            + ' to learn more about interactive tables.';\n",
              "          element.innerHTML = '';\n",
              "          dataTable['output_type'] = 'display_data';\n",
              "          await google.colab.output.renderOutput(dataTable, element);\n",
              "          const docLink = document.createElement('div');\n",
              "          docLink.innerHTML = docLinkHtml;\n",
              "          element.appendChild(docLink);\n",
              "        }\n",
              "      </script>\n",
              "    </div>\n",
              "  </div>\n",
              "  "
            ]
          },
          "metadata": {},
          "execution_count": 66
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "Here we have the data, but is it ready to train ?\n",
        "\n",
        "I don't think so, I think we need to first convert them into integers/numeric values so that our machine can learn.\n",
        "\n",
        "To do that, we will do numerical encoding using pandas, and pandas has a function called `get_dummies()`"
      ],
      "metadata": {
        "id": "ebUfcubGzm3M"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "# let's numerical encode all the data and store into a variable\n",
        "insurance_one_hot = pd.get_dummies(insurance)\n",
        "insurance_one_hot.head()"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 206
        },
        "id": "6I6AoIcW4F9K",
        "outputId": "2e308033-12f0-4514-f5da-46e9b95180e2"
      },
      "execution_count": 71,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "   age     bmi  children      charges  sex_female  sex_male  smoker_no  \\\n",
              "0   19  27.900         0  16884.92400           1         0          0   \n",
              "1   18  33.770         1   1725.55230           0         1          1   \n",
              "2   28  33.000         3   4449.46200           0         1          1   \n",
              "3   33  22.705         0  21984.47061           0         1          1   \n",
              "4   32  28.880         0   3866.85520           0         1          1   \n",
              "\n",
              "   smoker_yes  region_northeast  region_northwest  region_southeast  \\\n",
              "0           1                 0                 0                 0   \n",
              "1           0                 0                 0                 1   \n",
              "2           0                 0                 0                 1   \n",
              "3           0                 0                 1                 0   \n",
              "4           0                 0                 1                 0   \n",
              "\n",
              "   region_southwest  \n",
              "0                 1  \n",
              "1                 0  \n",
              "2                 0  \n",
              "3                 0  \n",
              "4                 0  "
            ],
            "text/html": [
              "\n",
              "  <div id=\"df-ebbd4fd2-0a74-47fe-86c4-919b905f28c4\">\n",
              "    <div class=\"colab-df-container\">\n",
              "      <div>\n",
              "<style scoped>\n",
              "    .dataframe tbody tr th:only-of-type {\n",
              "        vertical-align: middle;\n",
              "    }\n",
              "\n",
              "    .dataframe tbody tr th {\n",
              "        vertical-align: top;\n",
              "    }\n",
              "\n",
              "    .dataframe thead th {\n",
              "        text-align: right;\n",
              "    }\n",
              "</style>\n",
              "<table border=\"1\" class=\"dataframe\">\n",
              "  <thead>\n",
              "    <tr style=\"text-align: right;\">\n",
              "      <th></th>\n",
              "      <th>age</th>\n",
              "      <th>bmi</th>\n",
              "      <th>children</th>\n",
              "      <th>charges</th>\n",
              "      <th>sex_female</th>\n",
              "      <th>sex_male</th>\n",
              "      <th>smoker_no</th>\n",
              "      <th>smoker_yes</th>\n",
              "      <th>region_northeast</th>\n",
              "      <th>region_northwest</th>\n",
              "      <th>region_southeast</th>\n",
              "      <th>region_southwest</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>19</td>\n",
              "      <td>27.900</td>\n",
              "      <td>0</td>\n",
              "      <td>16884.92400</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>1</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>18</td>\n",
              "      <td>33.770</td>\n",
              "      <td>1</td>\n",
              "      <td>1725.55230</td>\n",
              "      <td>0</td>\n",
              "      <td>1</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>28</td>\n",
              "      <td>33.000</td>\n",
              "      <td>3</td>\n",
              "      <td>4449.46200</td>\n",
              "      <td>0</td>\n",
              "      <td>1</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3</th>\n",
              "      <td>33</td>\n",
              "      <td>22.705</td>\n",
              "      <td>0</td>\n",
              "      <td>21984.47061</td>\n",
              "      <td>0</td>\n",
              "      <td>1</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4</th>\n",
              "      <td>32</td>\n",
              "      <td>28.880</td>\n",
              "      <td>0</td>\n",
              "      <td>3866.85520</td>\n",
              "      <td>0</td>\n",
              "      <td>1</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>\n",
              "      <button class=\"colab-df-convert\" onclick=\"convertToInteractive('df-ebbd4fd2-0a74-47fe-86c4-919b905f28c4')\"\n",
              "              title=\"Convert this dataframe to an interactive table.\"\n",
              "              style=\"display:none;\">\n",
              "        \n",
              "  <svg xmlns=\"http://www.w3.org/2000/svg\" height=\"24px\"viewBox=\"0 0 24 24\"\n",
              "       width=\"24px\">\n",
              "    <path d=\"M0 0h24v24H0V0z\" fill=\"none\"/>\n",
              "    <path d=\"M18.56 5.44l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94zm-11 1L8.5 8.5l.94-2.06 2.06-.94-2.06-.94L8.5 2.5l-.94 2.06-2.06.94zm10 10l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94z\"/><path d=\"M17.41 7.96l-1.37-1.37c-.4-.4-.92-.59-1.43-.59-.52 0-1.04.2-1.43.59L10.3 9.45l-7.72 7.72c-.78.78-.78 2.05 0 2.83L4 21.41c.39.39.9.59 1.41.59.51 0 1.02-.2 1.41-.59l7.78-7.78 2.81-2.81c.8-.78.8-2.07 0-2.86zM5.41 20L4 18.59l7.72-7.72 1.47 1.35L5.41 20z\"/>\n",
              "  </svg>\n",
              "      </button>\n",
              "      \n",
              "  <style>\n",
              "    .colab-df-container {\n",
              "      display:flex;\n",
              "      flex-wrap:wrap;\n",
              "      gap: 12px;\n",
              "    }\n",
              "\n",
              "    .colab-df-convert {\n",
              "      background-color: #E8F0FE;\n",
              "      border: none;\n",
              "      border-radius: 50%;\n",
              "      cursor: pointer;\n",
              "      display: none;\n",
              "      fill: #1967D2;\n",
              "      height: 32px;\n",
              "      padding: 0 0 0 0;\n",
              "      width: 32px;\n",
              "    }\n",
              "\n",
              "    .colab-df-convert:hover {\n",
              "      background-color: #E2EBFA;\n",
              "      box-shadow: 0px 1px 2px rgba(60, 64, 67, 0.3), 0px 1px 3px 1px rgba(60, 64, 67, 0.15);\n",
              "      fill: #174EA6;\n",
              "    }\n",
              "\n",
              "    [theme=dark] .colab-df-convert {\n",
              "      background-color: #3B4455;\n",
              "      fill: #D2E3FC;\n",
              "    }\n",
              "\n",
              "    [theme=dark] .colab-df-convert:hover {\n",
              "      background-color: #434B5C;\n",
              "      box-shadow: 0px 1px 3px 1px rgba(0, 0, 0, 0.15);\n",
              "      filter: drop-shadow(0px 1px 2px rgba(0, 0, 0, 0.3));\n",
              "      fill: #FFFFFF;\n",
              "    }\n",
              "  </style>\n",
              "\n",
              "      <script>\n",
              "        const buttonEl =\n",
              "          document.querySelector('#df-ebbd4fd2-0a74-47fe-86c4-919b905f28c4 button.colab-df-convert');\n",
              "        buttonEl.style.display =\n",
              "          google.colab.kernel.accessAllowed ? 'block' : 'none';\n",
              "\n",
              "        async function convertToInteractive(key) {\n",
              "          const element = document.querySelector('#df-ebbd4fd2-0a74-47fe-86c4-919b905f28c4');\n",
              "          const dataTable =\n",
              "            await google.colab.kernel.invokeFunction('convertToInteractive',\n",
              "                                                     [key], {});\n",
              "          if (!dataTable) return;\n",
              "\n",
              "          const docLinkHtml = 'Like what you see? Visit the ' +\n",
              "            '<a target=\"_blank\" href=https://colab.research.google.com/notebooks/data_table.ipynb>data table notebook</a>'\n",
              "            + ' to learn more about interactive tables.';\n",
              "          element.innerHTML = '';\n",
              "          dataTable['output_type'] = 'display_data';\n",
              "          await google.colab.output.renderOutput(dataTable, element);\n",
              "          const docLink = document.createElement('div');\n",
              "          docLink.innerHTML = docLinkHtml;\n",
              "          element.appendChild(docLink);\n",
              "        }\n",
              "      </script>\n",
              "    </div>\n",
              "  </div>\n",
              "  "
            ]
          },
          "metadata": {},
          "execution_count": 71
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "# next thing is we have to separate the label and features into different variables\n",
        "X = insurance_one_hot.drop(\"charges\", axis = 1)\n",
        "y = insurance_one_hot[\"charges\"]"
      ],
      "metadata": {
        "id": "dROgmKYQ4ND1"
      },
      "execution_count": 72,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "# lets see the shape\n",
        "X.shape, y.shape"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "pvroMZdA4S8d",
        "outputId": "a6af6b05-632b-48fa-8d21-f347f81b72fe"
      },
      "execution_count": 73,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "((1338, 11), (1338,))"
            ]
          },
          "metadata": {},
          "execution_count": 73
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "Now I guess the data is ready to train into the model.\n",
        "\n",
        "But wait, we have to do one more thing, which is to create separate set of training and test set."
      ],
      "metadata": {
        "id": "8zflveap4zn2"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "# import sklearn function to do that\n",
        "from sklearn.model_selection import train_test_split\n",
        "\n",
        "X_train, X_test, y_train, y_test = train_test_split(X, y , test_size=0.2)"
      ],
      "metadata": {
        "id": "jJyyPJDc47ve"
      },
      "execution_count": 75,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "X_train.shape, X_test.shape"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "mcckZ_iq5RIV",
        "outputId": "b9188acc-0e4f-4f06-b806-df8161a4f2bf"
      },
      "execution_count": 76,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "((1070, 11), (268, 11))"
            ]
          },
          "metadata": {},
          "execution_count": 76
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "Yeah !! Now we are ready to go."
      ],
      "metadata": {
        "id": "hI1uaUQo5WW4"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "# first set the random seed\n",
        "tf.random.set_seed(42)\n",
        "\n",
        "# 1. Create a model\n",
        "model = tf.keras.Sequential([\n",
        "    tf.keras.layers.Dense(10),\n",
        "    tf.keras.layers.Dense(1)\n",
        "])\n",
        "\n",
        "# 2. Compile a model\n",
        "model.compile(loss=tf.keras.losses.mae,\n",
        "              optimizer= tf.keras.optimizers.SGD(),\n",
        "              metrics=[\"mae\"])\n",
        "\n",
        "# 3. Fit the model\n",
        "model.fit(X_train, y_train, epochs = 100, verbose = 0)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "f-8tmCz25bNA",
        "outputId": "15d9850a-a18f-4a9c-ae1e-f13585006e79"
      },
      "execution_count": 77,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "<keras.callbacks.History at 0x7fe5b101a890>"
            ]
          },
          "metadata": {},
          "execution_count": 77
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "# let's evaluate the model\n",
        "model.evaluate(X_test, y_test)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "V0xMc1Ak51bT",
        "outputId": "a6e3067a-d981-410f-d7b8-f55105f3dd64"
      },
      "execution_count": 78,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "9/9 [==============================] - 0s 2ms/step - loss: 8574.8662 - mae: 8574.8662\n"
          ]
        },
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "[8574.8662109375, 8574.8662109375]"
            ]
          },
          "metadata": {},
          "execution_count": 78
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "It says, that we 8575 ahead or behind the actual results, or loss. So we need to improve the model.\n",
        "\n",
        "How can we do this ? By following these steps,\n",
        "\n",
        "* Add extra dense layer and units(number of neurons)\n",
        "* Choose different optimizer function\n",
        "* Train a little longer, increase the number of epochs"
      ],
      "metadata": {
        "id": "AI13PusD58Ui"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "#Let's do the above then\n",
        "\n",
        "tf.random.set_seed(42)\n",
        "\n",
        "# create a model\n",
        "insurance_model = tf.keras.Sequential([\n",
        "    tf.keras.layers.Dense(100),\n",
        "    tf.keras.layers.Dense(10),\n",
        "    tf.keras.layers.Dense(1)\n",
        "])\n",
        "# compile a model\n",
        "insurance_model.compile(loss=tf.keras.losses.mae,\n",
        "                        optimizer=tf.keras.optimizers.Adam(),\n",
        "                        metrics=[\"mae\"])\n",
        "\n",
        "# fit the model\n",
        "history = insurance_model.fit(X_train, y_train, epochs = 200, verbose = 0)"
      ],
      "metadata": {
        "id": "8zQC8O-K6YaN"
      },
      "execution_count": 82,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "# evaluate the model\n",
        "insurance_model.evaluate(X_test, y_test)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "BqXVi7jl7CyW",
        "outputId": "9497338a-c45c-468a-b9db-f2d0a13e19e1"
      },
      "execution_count": 83,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "9/9 [==============================] - 0s 3ms/step - loss: 2880.7292 - mae: 2880.7292\n"
          ]
        },
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "[2880.729248046875, 2880.729248046875]"
            ]
          },
          "metadata": {},
          "execution_count": 83
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "Woah !! Much better"
      ],
      "metadata": {
        "id": "fnKe_lC07bMb"
      }
    },
    {
      "cell_type": "markdown",
      "source": [
        "Notice, we save the `fit()` into a `history` variable, just to plot the history of each epoch and training process, from `epoch` = 1 t0 200, what was the value of `Loss`"
      ],
      "metadata": {
        "id": "e-aPin297eQX"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "# plot the history (also known as loss curve)\n",
        "pd.DataFrame(history.history).plot()\n",
        "plt.ylabel(\"Loss\")\n",
        "plt.xlabel(\"Epochs\")"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 296
        },
        "id": "g-gdmJmu7vWN",
        "outputId": "3ef44c55-ee98-4889-9335-196f0cd9ca33"
      },
      "execution_count": 84,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "Text(0.5, 0, 'Epochs')"
            ]
          },
          "metadata": {},
          "execution_count": 84
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ],
            "image/png": "iVBORw0KGgoAAAANSUhEUgAAAZEAAAEGCAYAAACkQqisAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADh0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uMy4yLjIsIGh0dHA6Ly9tYXRwbG90bGliLm9yZy+WH4yJAAAgAElEQVR4nO3deXxc5Xn3/881i/bF8oKNNyyDscE2XpBtlrIEGrakkIQ8BEqLISxJHpo0pT8SSJ4+TZ+SkJCFhqRASFnbEJYQipuEnYQlAa94N7aFsbGMF8mLbGsdzVy/P+YYhGMZWZqZI3m+79drXjpzz5lzrjkazVf3OWfuY+6OiIhIT0TCLkBERPovhYiIiPSYQkRERHpMISIiIj2mEBERkR6LhV1Arg0ePNjHjBkTdhkiIv3KwoULG9x9yP7teRciY8aMYcGCBWGXISLSr5jZhgO1a3eWiIj0mEJERER6TCEiIiI9lnfHREREeiqRSFBXV0dra2vYpWRNUVERI0eOJB6Pd2t+hYiISDfV1dVRXl7OmDFjMLOwy8k4d2f79u3U1dVRXV3dredod5aISDe1trYyaNCgwzJAAMyMQYMGHVJPSyEiInIIDtcA2edQX59CpBtSySTznridRU/fH3YpIiJ9ikKkG8yMqlW/YMi820glk2GXIyJ5rKysLOwSPkQh0g0WidA49TpG+Xsse/nxsMsREekzFCLdNOWc2WxlENG5d4VdiogI7s6NN97IpEmTmDx5Mo8++igAmzdv5vTTT2fq1KlMmjSJV199lWQyyZVXXvn+vLfffnvG6tApvt0ULyjknbGXc9K6O9i4dgmjxk0JuyQRCdG//M8KVr63O6PLPH54Bf/8VxO7Ne+vf/1rFi9ezJIlS2hoaGDGjBmcfvrpPPzww5x77rl885vfJJlM0tzczOLFi9m0aRPLly8HYNeuXRmrWT2RQzD8pIsB2Lz85ZArEZF899prr3HZZZcRjUYZOnQoZ5xxBvPnz2fGjBncf//9fOtb32LZsmWUl5czduxY1q1bx5e//GWeeeYZKioqMlaHeiKHYOTRk9lNCV6nUYBF8l13ewy5dvrpp/PKK6/w29/+liuvvJIbbriBK664giVLlvDss89y991389hjj3HfffdlZH3qiRyCSDTKhsLxDNq1POxSRCTPnXbaaTz66KMkk0nq6+t55ZVXmDlzJhs2bGDo0KFce+21XHPNNSxatIiGhgZSqRQXX3wxt9xyC4sWLcpYHeqJHKK9g6dwXN1DtDbvpaikb51qJyL549Of/jSvv/46U6ZMwcy47bbbGDZsGA8++CDf//73icfjlJWV8dBDD7Fp0yauuuoqUqkUALfeemvG6jB3z9jC+oOamhrvzUWp3nzuv5j2p+t564JfMWHmxzNYmYj0datWreK4444Lu4ysO9DrNLOF7l6z/7zanXWIRk06DYBdtW+EXImISPgUIodo8PCj2MogYpvfDLsUEZHQKUR6YGvRGCpaNoZdhohI6BQiPdBaNJSqjvqwyxARCZ1CpAeS5cMZ5LtItLeFXYqISKgUIj0QqRhOxJztW7VLS0Tym0KkB4oGjQJg15b14RYiIhIyhUgPlB9xFABN9eqJiEh+U4j0wMBh6RBJ7KoLuRIRySfr169nwoQJXHnllRx77LFcfvnlvPDCC5x66qmMGzeOefPmMW/ePE4++WSmTZvGKaecwurVqwFIJpPceOONzJgxgxNOOIGf/exnGalJw570QOXAI2j1ODRuCrsUEQnL0zfBlmWZXeawyXD+dw86S21tLY8//jj33XcfM2bM4OGHH+a1115jzpw5fOc73+Ghhx7i1VdfJRaL8cILL/CNb3yDJ554gnvvvZfKykrmz59PW1sbp556Kueccw7V1dW9Klkh0gMWidAQGUy8eUvYpYhInqmurmby5MkATJw4kbPPPhszY/Lkyaxfv57GxkZmz57N2rVrMTMSiQQAzz33HEuXLuVXv/oVAI2Njaxdu1YhEpbG+GBKWreFXYaIhOUjegzZUlhY+P50JBJ5/34kEqGjo4N/+qd/4mMf+xhPPvkk69ev58wzzwTSV0L8yU9+wrnnnpvRenRMpIdaioZSqS8cikgf09jYyIgRIwB44IEH3m8/99xzueuuu97vmaxZs4ampqZer08h0kOJ0uEMTm0nlUyGXYqIyPu+9rWvcfPNNzNt2jQ6Ojreb7/mmms4/vjjmT59OpMmTeILX/jChx7vKQ0F30NzH7mVWW99l+1fWsGgoSMzUJmI9HUaCl5DwWdMvCodHDv1hUMRyWMKkR4qGXgkAE07NodciYhIeBQiPVRcMRiAxN4dIVciIrl0uB8CONTXl7UQMbP7zGybmS3v1PZ9M3vLzJaa2ZNmNqDTYzebWa2ZrTazczu1nxe01ZrZTZ3aq81sbtD+qJkVZOu1HEhp5UAAks27crlaEQlRUVER27dvP2yDxN3Zvn07RUVF3X5ONr8n8gDwU+ChTm3PAze7e4eZfQ+4Gfi6mR0PXApMBIYDL5jZscFz/h34OFAHzDezOe6+EvgecLu7P2JmdwNXA3dl8fV8SPmAdE8k1bIzV6sUkZCNHDmSuro66usP39P7i4qKGDmy+ycLZS1E3P0VMxuzX9tzne6+AXw2mL4IeMTd24B3zKwWmBk8Vuvu6wDM7BHgIjNbBZwF/HUwz4PAt8hhiBQWldDiBViLeiIi+SIej/f6G96HmzCPiXweeDqYHgF0HhK3Lmjrqn0QsMvdO/ZrPyAzu87MFpjZgkz+B7HHyoi0NWZseSIi/U0oIWJm3wQ6gF/kYn3ufo+717h7zZAhQzK23OZIGbHEnowtT0Skv8n52FlmdiXwSeBs/+Do1CZgVKfZRgZtdNG+HRhgZrGgN9J5/pxpiZZRmNid69WKiPQZOe2JmNl5wNeAC929udNDc4BLzazQzKqBccA8YD4wLjgTq4D0wfc5Qfj8ng+OqcwGnsrV69inLVZBUVI9ERHJX9k8xfeXwOvAeDOrM7OrSZ+tVQ48b2aLg7OqcPcVwGPASuAZ4Hp3Twa9jL8DngVWAY8F8wJ8HbghOAg/CLg3W6+lK4mCCkoUIiKSx7J5dtZlB2ju8oPe3b8NfPsA7b8DfneA9nV8cAZXKJIFlZTS+1EwRUT6K31jvRe8aAAVNJPMwEiYIiL9kUKkF6y4EoC9jdtDrkREJBwKkV6IllQBsLexIeRKRETCoRDphXhpevysZvVERCRPKUR6oaA8HSKtezR+lojkJ4VILxRXDAKgXcPBi0ieUoj0QklFuifS0aQQEZH8pBDphQ+Gg9dIviKSnxQivVBcUk67R6FVISIi+Ukh0gsWiWg4eBHJawqRXmqKlBFTiIhInlKI9FJLpIwCDQcvInlKIdJLbbEKCpN7wy5DRCQUCpFeSsRKKUppJF8RyU8KkV5KxssoSrWEXYaISCgUIr2UKiij9EMXaRQRyR8KkV7ygnJKrZVUMhl2KSIiOacQ6SUrKgegaa9O8xWR/KMQ6aVIYTpEmjWSr4jkIYVIL0WLKwBoVU9ERPKQQqSXYu+HiMbPEpH8oxDppYLS9HXW25vVExGR/KMQ6aXC0gEAdLRo6BMRyT8KkV4qCnoiiWaFiIjkH4VIL5WUp3siqVaFiIjkH4VIL+0LEW/bE3IlIiK5pxDppYLCIto8DgoREclDCpEMaLJiIu0aDl5E8o9CJAOarYRoQiEiIvlHIZIBbZESYh26poiI5B+FSAa0RUuJK0REJA8pRDKgPVpKYVIhIiL5J2shYmb3mdk2M1veqW2gmT1vZmuDn1VBu5nZHWZWa2ZLzWx6p+fMDuZfa2azO7WfaGbLgufcYWaWrdfyUTpipRSldGEqEck/2eyJPACct1/bTcCL7j4OeDG4D3A+MC64XQfcBenQAf4ZmAXMBP55X/AE81zb6Xn7rytnkgVlFLkukSsi+SdrIeLurwA79mu+CHgwmH4Q+FSn9oc87Q1ggJkdCZwLPO/uO9x9J/A8cF7wWIW7v+HuDjzUaVk553FdIldE8lOuj4kMdffNwfQWYGgwPQLY2Gm+uqDtYO11B2g/IDO7zswWmNmC+vr63r2CA/DCcoqtnY5Ee8aXLSLSl4V2YD3oQXiO1nWPu9e4e82QIUMyvnwLrm7YtEfDwYtIfsl1iGwNdkUR/NwWtG8CRnWab2TQdrD2kQdoD0W0aN8lcvffeycicnjLdYjMAfadYTUbeKpT+xXBWVonAY3Bbq9ngXPMrCo4oH4O8Gzw2G4zOyk4K+uKTsvKuaiubigieSqWrQWb2S+BM4HBZlZH+iyr7wKPmdnVwAbgkmD23wEXALVAM3AVgLvvMLN/BeYH8/0/d9/37/7/Jn0GWDHwdHALRbwkfU2RtiYNBy8i+SVrIeLul3Xx0NkHmNeB67tYzn3AfQdoXwBM6k2NmRIvSfdE2pt0TERE8ou+sZ4BRWXpa4okmrU7S0Tyi0IkA0oqBgHQ0ayeiIjkF4VIBpRWpL9E760KERHJLwqRDCgpraDDI7iusy4ieUYhkgEWibDXSoi0qSciIvlFIZIhzVZKtF3XWReR/KIQyZCWSCmxDl0iV0Tyi0IkQ1qjZRQk1BMRkfyiEMmQ9lgZRUn1REQkvyhEMqQjXk5xSpfIFZH8ohDJkFRBOaUoREQkvyhEMsQLKynzFlLJZNiliIjkjEIkU4oqiJjTtFffFRGR/KEQyZBIcXo4+KbG7SFXIiKSO90KETMrNbNIMH2smV1oZvHslta/xErSI/m26OqGIpJHutsTeQUoMrMRwHPA35K+IJQE4qX7QmRnyJWIiOROd0PE3L0Z+Axwp7v/L2Bi9srqfwrL0iP5tjfpmiIikj+6HSJmdjJwOfDboC2anZL6p+LydIh0KEREJI90N0S+CtwMPOnuK8xsLPD77JXV/xSXDwQg2aIQEZH80a1rrLv7y8DLAMEB9gZ3/0o2C+tvyirTIeItuqaIiOSP7p6d9bCZVZhZKbAcWGlmN2a3tP6lqLiUNo/juqaIiOSR7u7OOt7ddwOfAp4GqkmfoSWdpC9MpZ6IiOSP7oZIPPheyKeAOe6eADx7ZfVPzVZKTMPBi0ge6W6I/AxYD5QCr5jZUYD+5d5PS7SMeELDwYtI/ujugfU7gDs6NW0ws49lp6T+qy1aSmGHeiIikj+6e2C90sx+ZGYLgtsPSfdKpJNEvJzilEJERPJHd3dn3QfsAS4JbruB+7NVVH/VXjKMwckGPJUKuxQRkZzo1u4s4Gh3v7jT/X8xs8XZKKhfG3AUJdva2NGwmYFHjAi7GhGRrOtuT6TFzP5i3x0zOxVoyU5J/Vfh4DEANNTVhluIiEiOdLcn8kXgITOrDO7vBGZnp6T+q3L4OAD2bFkHnBFuMSIiOdDds7OWAFPMrCK4v9vMvgoszWZx/c2QUekQSTSsC7kSEZHcOKQrG7r77uCb6wA3ZKGefq28ciC7KMMa3w27FBGRnOjN5XGtx080+wczW2Fmy83sl2ZWZGbVZjbXzGrN7FEzKwjmLQzu1waPj+m0nJuD9tVmdm4vXkvGNESHUty0KewyRERyojch0qNhT4KrI34FqHH3SaSvS3Ip8D3gdnc/hvQxl6uDp1wN7Azabw/mw8yOD543ETgPuNPMQr/Gye6i4VS2bw67DBGRnDhoiJjZHjPbfYDbHmB4L9YbA4rNLAaUAJuBs4BfBY8/SHqcLoCLgvsEj59tZha0P+Lube7+DlALzOxFTRnRXjaSocmt+q6IiOSFg4aIu5e7e8UBbuXu3t0zu/Zf5ibgB8C7pMOjEVgI7HL3jmC2OmDfFy1GABuD53YE8w/q3H6A53yImV2379v29fX1PSm726zqKIoswfZtdVldj4hIX9Cb3Vk9YmZVpHsR1aR7M6Wkd0dljbvf4+417l4zZMiQbK6KoiOqAWioW5vV9YiI9AU5DxHgL4F33L0+GFL+18CpwIBg9xbASGDf0elNwCiA4PFKYHvn9gM8JzQDjkyf5tu4blHIlYiIZF8YIfIucJKZlQTHNs4GVpK+Zvtng3lmA08F03P44IuNnwVecncP2i8Nzt6qBsYB83L0Gro06tiprI0ew5iVd9G8V1c5FJHDW85DxN3nkj5AvghYFtRwD/B14AYzqyV9zOPe4Cn3AoOC9huAm4LlrAAeIx1AzwDXu3syhy/lgCLRKMlzbmUo21n60I20NGlUXxE5fFn6n/r8UVNT4wsWLMj6eubffgkzGp+lxQtYXTKdtrHnMPbUixkyfEzW1y0ikmlmttDda/6sXSGSHR2Jdla9/lual/2GUQ2vMNy3AbA2egwNI85iyIkXcfTkU7BIGHsURUQOjUIkkKsQ6cxTKda/tZAt85+kauNLHJt4i4g52xjIO4NOo2jiBYw/6ZMUlZTltC4Rke5SiATCCJH9bd9ax9t/epJY7bOM3zufUmulxQt4q7SGxNHnMPbkzzB4+FGh1igi0plCJNAXQqSzttZmVr/xNC3Lf8NRDa8yjPSXIdfGxtEw/CyGnfS/qD5+RshViki+U4gE+lqIdOapFO+snM/WBU8ysO73jEusJmLO29Fq6sdcSPXHZjN05NFhlykieUghEujLIbK/hi0befsP/8WA2icZ37GalBurCk+gecLFjD/rb6gYMCjsEkUkTyhEAv0pRDrbWLuMulceYtTG/2Gkb6bN46woPxk74XNMPOOzFBQWhV2iiBzGFCKB/hoi+3gqxZpFf2DX3F8wrv45BrKbnVTw1pGf4qhzrmd49YSwSxSRw5BCJNDfQ6SzRHsbK197itTCBzlh7x8xYFnJTJhxDZPPuJhINPTLq4jIYUIhEjicQqSzLRtrWf/snRxT9wSD2cV7NpR3x17K+PO+RNWQI8MuT0T6OYVI4HANkX3a21pZ9sJ/UbzkAY5vX0abx1k64CwqTvsi42vOCrs8EemnFCKBwz1EOntn5Xy2vXQnk+p/R6m1sjo2gb3Tr2PKx/+WWLwg7PJEpB9RiATyKUT22bt7J8t/excj1zzESN/MZoawqeZrnHjBNRq7S0S6RSESyMcQ2SfZ0cGy3z9G2Rs/4Jjk26yNjWPvzK8y5ezLdBBeRA6qqxDRv6F5JBqLMfXjf031zfOYN+VfKUnuZtqfrufdb09lwZy76Ui0h12iiPQzCpE8FI3FmPnprzD0G8tZcOJtANQs+jpbvzOJuY//kLbW5pArFJH+QruzhFQyyZIXf0npvB9zbMcatjGQdeOu5Ljz/zeVA4eEXZ6I9AE6JhJQiHTNUymWv/Y/RF77IRPbl9DmcZaXn0ps+uVMOOWTFBaVhF2iiIREIRJQiHRP7ZI/0vDa/Yyvf4Yq9tDshdQWn0DzqNMYcOypHHX8LIpLy8MuU0RyRCESUIgcmva2Vla++iRtbz3PsB1zOSpVB0DSjY3RkewsGk1b+WhiI6Yw5NhZjDzmBKKxWMhVi0imKUQCCpHe2bbpHd5b+Tot7y6kePsKKlvfY1hyM8WWPrOr2QvZUHAMjVUTiY6YxuBxMxg1boq+3CjSzylEAgqRzOtItLNx7RIa1swltelNKneu4KjE2+8HS6vHeTdezc7K42DYCVQdXcPoCTW6prxIP6IQCShEcuP9YFk7n+R7SyjfuZJR7bVUkD59uMMj1EVH0lA2no6hk6kYO5Mxk0+hpKwy5MpF5EAUIgGFSHg8lWLzhjVsXTOX1o2LKdm+giNb1nIEO4B0sKyPVbO9agqRUTMYNvE0Ro6dqKFZRPoAhUhAIdL3NGzZyKaVf6T57Tcob3iT6ta3KLVWAHZRxoai42geVkPlhDMZO+U0iopLQ65YJP8oRAIKkb4v2dHBu2vepH7Va1A3nyMalzEm9S4AbR7n7YLxNB4xg9Jxp3H0iWdTWj4g5IpFDn8KkYBCpH/a1bCF9W++RGvtKwzcvpCxiVpilqLdo6wtnMTuEacxZOoFjJ10kgaTFMkChUhAIXJ4aNqzi3WLXqJp1Qscse2PjE2tB2A7lbxTMRM/+iyqZ/0Vg4eNCrdQkcOEQiSgEDk8Nby3gXfm/QZb9xJH755HFbsBeDs6lm1H/AUVk85jXM3ZFBQWhVypSP+kEAkoRA5/qWSSdctfp37x76jY9ArHtq0kbkmavIg1pdNpH3MWE/7ySg0uKXIIFCIBhUj+2dO4g9q5v6N99fOM3v5HjqSeZi9k2ZBPUHXKbMZNPV2nEYt8BIVIQCGS3zyVYt3yN9jx4r9xwq6XKLQEG204daMv5KiPfZ7hY8aHXaJIn9SnrmxoZgPM7Fdm9paZrTKzk81soJk9b2Zrg59VwbxmZneYWa2ZLTWz6Z2WMzuYf62ZzQ7jtUj/YpEIR59wCjP+4THa/mE18yb/C7vjgzh5w90Mf2AmK799KvN+9SMadzaEXapIvxBKT8TMHgRedff/MLMCoAT4BrDD3b9rZjcBVe7+dTO7APgycAEwC/ixu88ys4HAAqAGcGAhcKK77zzYutUTkQN5b/1qNvzhfka8O4fRqU20eZxllWdSceaXOXb6GWGXJxK6PrM7y8wqgcXAWO+0cjNbDZzp7pvN7EjgD+4+3sx+Fkz/svN8+27u/oWg/UPzdUUhIgfjqRRrF7/Kzj89wMT6pymzFlbHxrNnytWccM5snd0leasv7c6qBuqB+83sTTP7DzMrBYa6++Zgni3A0GB6BLCx0/Prgrau2v+MmV1nZgvMbEF9fX0GX4ocbiwS4djpZzDr7+6Hf1zF3Ak3UZLcQ83Cr9F46wTmPvZ9Eu1tYZcp0meEESIxYDpwl7tPA5qAmzrPEPRQMtZFcvd73L3G3WuGDNFpndI9ZRVVzLr0Zkb8n+UsPeNeGuLDmbXyFrbcOpU3n30QT6XCLlEkdGGESB1Q5+5zg/u/Ih0qW4PdWAQ/twWPbwI6f+14ZNDWVbtIRkWiUU742GeZcPNrLD71LlIWZdrrX2H1d05h1dxnwy5PJFQ5DxF33wJsNLN951KeDawE5gD7zrCaDTwVTM8BrgjO0joJaAx2ez0LnGNmVcGZXOcEbSJZYZEIUz/+14y4eRHzJn+LgR1bOe7pS3jztvNZs+jlsMsTCUVYZ2dNBf4DKADWAVeRDrTHgNHABuASd99hZgb8FDgPaAaucvcFwXI+T/qsLoBvu/v9H7VuHViXTGlp2sOSx2/l+PX3U0EzS4pnMvxvf86Q4WPCLk0k4/rM2VlhU4hIpu1p3MGKp25nytt302YFrJt1C9PPvyrsskQyqi+dnSVyWCmvHMhJV/wr9Zc/T33sSKbP/SoLfnQxW95dG3ZpIlmnEBHJkNHHTmXM1/7I66OuZXLjy1TdezJzH7st7LJEskohIpJB8YJCTr76B+y8+nXeKpnGrJXfZu5PZtPa0hR2aSJZoRARyYJho8cx6R+f5o1hlzNr+3+z+funsGH14rDLEsk4hYhIlkRjMU764p0sOf3nVKZ2MPDh81j84iNhlyWSUQoRkSybctYltH/+JbbGhjP11S/wxp3XaveWHDYUIiI5MGz0OEb+4yvMHfJZTtr2GG/ffj7NexvDLkuk1xQiIjlSVFLGrOvvZcH07zGhbSnv3PFJBYn0ewoRkRyrufCLvDnjNia0LWP9HZ+gac+usEsS6TGFiEgIaj55HW/O/D7j25azQUEi/ZhCRCQkNZ+4lsUzf8Cx7St5944L2Lv7oBflFOmTFCIiITrxE9ewZNaPGNe+irqfXMCexh1hlyRySBQiIiE78YKrWHrSjzim/S02KUikn1GIiPQB08+/iqUn387RiTW895Pz2b1re9gliXSLQkSkj5h+3pUsO+XHjE2sZfNPFSTSPyhERPqQ6ef+LctPvYOxiVrq7ryIRHtb2CWJHJRCRKSPmXbO37DkxO9wfPsyFt3zpbDLETkohYhIH1Rz4Rd5Y+hlzGp4gnlP3hF2OSJdUoiI9FE119zB8sKpTF38L6xZ9IewyxE5IIWISB8Vixcw4ppHaIgMZMCcq2jY8m7YJYn8GYWISB9WNeRIWj7zn5T7Xurv/Rztba1hlyTyIQoRkT7u6MknsWLmdzgusZI37/lC2OWIfIhCRKQfqPnEtbx+5N8wa/t/M++J28MuR+R9ChGRfmLm1T9madGJTF16C28teDHsckQAhYhIvxGNxTjq2l9SHxnEoN9cTcN7G8IuSUQhItKfVA4aSvtn/5NSb6bhvs/R1tocdkmS5xQiIv1M9cRZvHXS95jQsYrFOtAuIVOIiPRD08+/itdHXMmsHXOY+/gPwi5H8phCRKSfmnnVD1lSNINpy7/Dij/9LuxyJE8pRET6qWgsxpgvPMLm6HBGP/t53l76p7BLkjykEBHpxyqrBlP0+adosjIG/voSVr7xTNglSZ5RiIj0c0NHHk3yiqfYE6ngmKf/mtd//lV21m8OuyzJE6GFiJlFzexNM/tNcL/azOaaWa2ZPWpmBUF7YXC/Nnh8TKdl3By0rzazc8N5JSLhGzF2IpV/9zLLKs5gVt0DxH86jbmPfpdUMhl2aXKYC7Mn8vfAqk73vwfc7u7HADuBq4P2q4GdQfvtwXyY2fHApcBE4DzgTjOL5qh2kT6ncuAQTvzHJ9l42Uu8UzSBWatuZdMtk3jjF/+P9asWKFAkK8zdc79Ss5HAg8C3gRuAvwLqgWHu3mFmJwPfcvdzzezZYPp1M4sBW4AhwE0A7n5rsMz35zvYumtqanzBggXZemkifYKnUrz57IOULryL8R2rAWj1OJujI9hZPJq2ymqiQ8ZRMfI4hlVPYsDgYaSSScwMi2gvt/w5M1vo7jX7t8fCKAb4N+BrQHlwfxCwy907gvt1wIhgegSwESAImMZg/hHAG52W2fk5H2Jm1wHXAYwePTpzr0Kkj7JIhOnnXwXnX8WmdavYtOQFUltXUbR7HYNb3ubIvX8k/l4SlqTnb/ZCCmmnlQI2x4bTFi0jESkiGSsmGS0hWVSFVR2FJ9shEqN89FQKywa8v754YTGxgkISbS0UFJURiUbZsmYBscISho+vYcCgYUSi2lFwOMp5iJjZJ4Ft7r7QzM7MxTrd/R7gHkj3RHKxTpG+YsTY4xgx9rgPtXUk2tm4YTU73l1By+Y10FiHx0uwRBNFe94lnmyipGMXBe1bKFvd+E4AAApBSURBVEy1MmDXLoq2Jj5YwOqPXu+wfRMvQodH2Gnl7I2U02FxUhYjSYxUJEbSYqQsRioSxy2GR9LtbjE8Gofgp0diEC2ASByiMSwah0gci8aJFlcSLxtAqqODaLyIgvIqorECzCJEojEsEiEWLyAaL2L31g04zujjZ1FaPuCAtXsqRSqVIhqLkWhvI5nsoKi4tGe/gMNcGD2RU4ELzewCoAioAH4MDDCzWNAbGQlsCubfBIwC6oLdWZXA9k7t+3R+jogcRCxewKhjJjPqmMndmj+VTNJQ/x4FRSW0N+9l89qFdLR9MG5XMtGCJ9qIxItItjfjiTYqx0wl0bqXproVeFMDkdYdxNt2YakEkVSCiHcQ8Q7iqTYi3kTUOz64kXz/Z5wEMU+Sjp0Oota7/wOH75t4+sPtLV5AwmIUeIIiSxAFkm7EzYkD79kRtFsRUe9gd3wIrQVVpKIFxBJNmCdxi1DethXHaCwdQ0fRQCgoh8IyiMbx1t3Ed71DonQokaoxeEcrFi8mVlpFQWkVBaUDKCguxR08lQSLUFhcxt7t77FnSy2FlUOpHFbNoOHVRKMx4gWFxOIFXf6+mvY2YmaUVVT1ant9lFCOiby/8nRP5P9z90+a2ePAE+7+iJndDSx19zvN7Hpgsrt/0cwuBT7j7peY2UTgYWAm6ffFi8A4dz/o0UMdExHp31LJJIlEGx2JdjoSCRLtrbTs2UFz43ai8QI62ltp27sTT3akP4w9me5ZJNpIJVopHDgCTyVp3bgUT7YBYO7Q0QrJ9nRvJ1aER6Lp+7EicKdg51oslcAtSmnbNkqSjcQ8QZuVkLQoUU+yp2AI5kkGt22klCbKvIVIEHopN+ptIFW+iwLLzEkOe72YAtpJEWFHZCBR76DEmyml9f311lNFuxUQ8w4GfmMV8YLCHq2rrx0TOZCvA4+Y2S3Am8C9Qfu9wH+aWS2wg/QZWbj7CjN7DFgJdADXf1SAiEj/F4lGKYyWUFhU8kHjsFFdP6FLl2aspq6kkkn2Nu0mmUxSWFTM0OJS2tta2dawmcKiEtpam2lqbKBl93YSzbtJtjWlT2ywKHiSjtYm4mUDqRo5geZdW2na+g4dje+Be7on09qIxwqxZIJYSz2pSAGpgnK8sBwrLMeTCWI7ajHvIBUpYEAq8x+RofZEwqCeiIjIoeuqJ6Jz+UREpMcUIiIi0mMKERER6TGFiIiI9JhCREREekwhIiIiPaYQERGRHlOIiIhIj+Xdlw3NrB7Y0MOnDwYaMlhOpqiuQ9dXa1Ndh6av1gV9t7ae1nWUuw/ZvzHvQqQ3zGzBgb6xGTbVdej6am2q69D01bqg79aW6bq0O0tERHpMISIiIj2mEDk094RdQBdU16Hrq7WprkPTV+uCvltbRuvSMREREekx9URERKTHFCIiItJjCpFuMLPzzGy1mdWa2U0h1zLKzH5vZivNbIWZ/X3Q/i0z22Rmi4PbBSHUtt7MlgXrXxC0DTSz581sbfAzuxd8/vOaxnfaJovNbLeZfTWs7WVm95nZNjNb3qntgNvI0u4I3ndLzWx6juv6vpm9Faz7STMbELSPMbOWTtvu7hzX1eXvzsxuDrbXajM7N8d1PdqppvVmtjhoz+X26urzIXvvMXfX7SA3IAq8DYwFCoAlwPEh1nMkMD2YLgfWAMcD3yJ9vfowt9V6YPB+bbcBNwXTNwHfC/l3uQU4KqztBZwOTAeWf9Q2Ai4AngYMOAmYm+O6zgFiwfT3OtU1pvN8IWyvA/7ugr+DJUAhUB383UZzVdd+j/8Q+L8hbK+uPh+y9h5TT+SjzQRq3X2du7cDjwAXhVWMu29290XB9B5gFTAirHq64SLgwWD6QeBTIdZyNvC2u/d0xIJec/dXgB37NXe1jS4CHvK0N4ABZnZkrupy9+fcvSO4+wYwMhvrPtS6DuIi4BF3b3P3d4Ba0n+/Oa3LzAy4BPhlNtZ9MAf5fMjae0wh8tFGABs73a+jj3xom9kYYBowN2j6u6BLel+udxsFHHjOzBaa2XVB21B33xxMbwGGhlDXPpfy4T/ssLfXPl1to7703vs86f9Y96k2szfN7GUzOy2Eeg70u+sr2+s0YKu7r+3UlvPttd/nQ9beYwqRfsrMyoAngK+6+27gLuBoYCqwmXR3Otf+wt2nA+cD15vZ6Z0f9HT/OZRzys2sALgQeDxo6gvb68+EuY26YmbfBDqAXwRNm4HR7j4NuAF42MwqclhSn/zddXIZH/5nJefb6wCfD+/L9HtMIfLRNgGjOt0fGbSFxszipN8gv3D3XwO4+1Z3T7p7Cvg5WerGH4y7bwp+bgOeDGrYuq97HPzcluu6AucDi9x9a1Bj6Nurk662UejvPTO7EvgkcHnw4UOwu2h7ML2Q9LGHY3NV00F+d31he8WAzwCP7mvL9fY60OcDWXyPKUQ+2nxgnJlVB//NXgrMCauYYH/rvcAqd/9Rp/bO+zE/DSzf/7lZrqvUzMr3TZM+KLuc9LaaHcw2G3gql3V18qH/DsPeXvvpahvNAa4IzqA5CWjstEsi68zsPOBrwIXu3typfYiZRYPpscA4YF0O6+rqdzcHuNTMCs2sOqhrXq7qCvwl8Ja71+1ryOX26urzgWy+x3JxxkB/v5E+g2EN6f8gvhlyLX9Buiu6FFgc3C4A/hNYFrTPAY7McV1jSZ8ZswRYsW87AYOAF4G1wAvAwBC2WSmwHajs1BbK9iIdZJuBBOn9z1d3tY1InzHz78H7bhlQk+O6aknvL9/3Prs7mPfi4He8GFgE/FWO6+rydwd8M9heq4Hzc1lX0P4A8MX95s3l9urq8yFr7zENeyIiIj2m3VkiItJjChEREekxhYiIiPSYQkRERHpMISIiIj2mEBHJADNL2odHC87YaM/BKLBhfo9FpEuxsAsQOUy0uPvUsIsQyTX1RESyKLiuxG2Wvs7KPDM7JmgfY2YvBYMIvmhmo4P2oZa+dseS4HZKsKiomf08uEbEc2ZWHMz/leDaEUvN7JGQXqbkMYWISGYU77c763OdHmt098nAT4F/C9p+Ajzo7ieQHtjwjqD9DuBld59C+noVK4L2ccC/u/tEYBfpb0FD+toQ04LlfDFbL06kK/rGukgGmNledy87QPt64Cx3XxcMjLfF3QeZWQPp4ToSQftmdx9sZvXASHdv67SMMcDz7j4uuP91IO7ut5jZM8Be4L+B/3b3vVl+qSIfop6ISPZ5F9OHoq3TdJIPjmd+gvTYR9OB+cEosiI5oxARyb7Pdfr5ejD9J9IjQgNcDrwaTL8IfAnAzKJmVtnVQs0sAoxy998DXwcqgT/rDYlkk/5rEcmMYjNb3On+M+6+7zTfKjNbSro3cVnQ9mXgfjO7EagHrgra/x64x8yuJt3j+BLp0WIPJAr8VxA0Btzh7rsy9opEukHHRESyKDgmUuPuDWHXIpIN2p0lIiI9pp6IiIj0mHoiIiLSYwoRERHpMYWIiIj0mEJERER6TCEiIiI99v8DkeFCbrwBGyMAAAAASUVORK5CYII=\n"
          },
          "metadata": {
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "In our case, the metrics that we used 'Mae' and loss is same, that's why the line is overlapped, but if we choose different in both then graph would be different.\n"
      ],
      "metadata": {
        "id": "_q7rctr38C2O"
      }
    },
    {
      "cell_type": "markdown",
      "source": [
        "> 🤔 **Question**: How long should we train ?  \n",
        "      It depends on our problem we're working on. Sometimes training won't take very long, other times it'll take longer than we expect. A common method is to set our model training for a very long time (e.g 1000s of epochs) but set it up with an EarlyStopping callback so it stops automatically when it stops improving."
      ],
      "metadata": {
        "id": "o2bWzcTi8YPA"
      }
    },
    {
      "cell_type": "markdown",
      "source": [
        "## Preprocessing data (normalization and standardization)\n",
        "\n",
        "Read more in the Resources section of the repo."
      ],
      "metadata": {
        "id": "iv0kMBOj9F-P"
      }
    },
    {
      "cell_type": "markdown",
      "source": [
        "The idea of this is to keep all the features in our data set in same scale from 0 to 1.\n",
        "\n",
        "Neural Network by default suggest that to use normalization technique over standardization to convert the features into same scale.\n",
        "\n",
        "Now, just as before, we need to transform the non-numerical columns into numbers and this time we'll also be normalizing the numerical columns with different ranges(0 to 1).\n",
        "\n",
        "To do this, we will use a few classed from Scikit-learn:\n",
        "\n",
        "* make_column_transformer - build a multi-step data Preprocessing function for the following transformations:\n",
        "  * `MinMaxScaler`: make sure all the columns are in same scale\n",
        "  * `OneHotEncoder`: one hot encode the non-numerical columns."
      ],
      "metadata": {
        "id": "RH3cqnZa9We1"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "insurance.head()"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 206
        },
        "id": "2Y6ERabD_YPL",
        "outputId": "1ae49424-b23e-427c-ff32-12de205222bc"
      },
      "execution_count": 87,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "   age     sex     bmi  children smoker     region      charges\n",
              "0   19  female  27.900         0    yes  southwest  16884.92400\n",
              "1   18    male  33.770         1     no  southeast   1725.55230\n",
              "2   28    male  33.000         3     no  southeast   4449.46200\n",
              "3   33    male  22.705         0     no  northwest  21984.47061\n",
              "4   32    male  28.880         0     no  northwest   3866.85520"
            ],
            "text/html": [
              "\n",
              "  <div id=\"df-536ddee2-1df0-4ee5-95f9-118a9e1cb99b\">\n",
              "    <div class=\"colab-df-container\">\n",
              "      <div>\n",
              "<style scoped>\n",
              "    .dataframe tbody tr th:only-of-type {\n",
              "        vertical-align: middle;\n",
              "    }\n",
              "\n",
              "    .dataframe tbody tr th {\n",
              "        vertical-align: top;\n",
              "    }\n",
              "\n",
              "    .dataframe thead th {\n",
              "        text-align: right;\n",
              "    }\n",
              "</style>\n",
              "<table border=\"1\" class=\"dataframe\">\n",
              "  <thead>\n",
              "    <tr style=\"text-align: right;\">\n",
              "      <th></th>\n",
              "      <th>age</th>\n",
              "      <th>sex</th>\n",
              "      <th>bmi</th>\n",
              "      <th>children</th>\n",
              "      <th>smoker</th>\n",
              "      <th>region</th>\n",
              "      <th>charges</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>19</td>\n",
              "      <td>female</td>\n",
              "      <td>27.900</td>\n",
              "      <td>0</td>\n",
              "      <td>yes</td>\n",
              "      <td>southwest</td>\n",
              "      <td>16884.92400</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>18</td>\n",
              "      <td>male</td>\n",
              "      <td>33.770</td>\n",
              "      <td>1</td>\n",
              "      <td>no</td>\n",
              "      <td>southeast</td>\n",
              "      <td>1725.55230</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>28</td>\n",
              "      <td>male</td>\n",
              "      <td>33.000</td>\n",
              "      <td>3</td>\n",
              "      <td>no</td>\n",
              "      <td>southeast</td>\n",
              "      <td>4449.46200</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3</th>\n",
              "      <td>33</td>\n",
              "      <td>male</td>\n",
              "      <td>22.705</td>\n",
              "      <td>0</td>\n",
              "      <td>no</td>\n",
              "      <td>northwest</td>\n",
              "      <td>21984.47061</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4</th>\n",
              "      <td>32</td>\n",
              "      <td>male</td>\n",
              "      <td>28.880</td>\n",
              "      <td>0</td>\n",
              "      <td>no</td>\n",
              "      <td>northwest</td>\n",
              "      <td>3866.85520</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>\n",
              "      <button class=\"colab-df-convert\" onclick=\"convertToInteractive('df-536ddee2-1df0-4ee5-95f9-118a9e1cb99b')\"\n",
              "              title=\"Convert this dataframe to an interactive table.\"\n",
              "              style=\"display:none;\">\n",
              "        \n",
              "  <svg xmlns=\"http://www.w3.org/2000/svg\" height=\"24px\"viewBox=\"0 0 24 24\"\n",
              "       width=\"24px\">\n",
              "    <path d=\"M0 0h24v24H0V0z\" fill=\"none\"/>\n",
              "    <path d=\"M18.56 5.44l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94zm-11 1L8.5 8.5l.94-2.06 2.06-.94-2.06-.94L8.5 2.5l-.94 2.06-2.06.94zm10 10l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94z\"/><path d=\"M17.41 7.96l-1.37-1.37c-.4-.4-.92-.59-1.43-.59-.52 0-1.04.2-1.43.59L10.3 9.45l-7.72 7.72c-.78.78-.78 2.05 0 2.83L4 21.41c.39.39.9.59 1.41.59.51 0 1.02-.2 1.41-.59l7.78-7.78 2.81-2.81c.8-.78.8-2.07 0-2.86zM5.41 20L4 18.59l7.72-7.72 1.47 1.35L5.41 20z\"/>\n",
              "  </svg>\n",
              "      </button>\n",
              "      \n",
              "  <style>\n",
              "    .colab-df-container {\n",
              "      display:flex;\n",
              "      flex-wrap:wrap;\n",
              "      gap: 12px;\n",
              "    }\n",
              "\n",
              "    .colab-df-convert {\n",
              "      background-color: #E8F0FE;\n",
              "      border: none;\n",
              "      border-radius: 50%;\n",
              "      cursor: pointer;\n",
              "      display: none;\n",
              "      fill: #1967D2;\n",
              "      height: 32px;\n",
              "      padding: 0 0 0 0;\n",
              "      width: 32px;\n",
              "    }\n",
              "\n",
              "    .colab-df-convert:hover {\n",
              "      background-color: #E2EBFA;\n",
              "      box-shadow: 0px 1px 2px rgba(60, 64, 67, 0.3), 0px 1px 3px 1px rgba(60, 64, 67, 0.15);\n",
              "      fill: #174EA6;\n",
              "    }\n",
              "\n",
              "    [theme=dark] .colab-df-convert {\n",
              "      background-color: #3B4455;\n",
              "      fill: #D2E3FC;\n",
              "    }\n",
              "\n",
              "    [theme=dark] .colab-df-convert:hover {\n",
              "      background-color: #434B5C;\n",
              "      box-shadow: 0px 1px 3px 1px rgba(0, 0, 0, 0.15);\n",
              "      filter: drop-shadow(0px 1px 2px rgba(0, 0, 0, 0.3));\n",
              "      fill: #FFFFFF;\n",
              "    }\n",
              "  </style>\n",
              "\n",
              "      <script>\n",
              "        const buttonEl =\n",
              "          document.querySelector('#df-536ddee2-1df0-4ee5-95f9-118a9e1cb99b button.colab-df-convert');\n",
              "        buttonEl.style.display =\n",
              "          google.colab.kernel.accessAllowed ? 'block' : 'none';\n",
              "\n",
              "        async function convertToInteractive(key) {\n",
              "          const element = document.querySelector('#df-536ddee2-1df0-4ee5-95f9-118a9e1cb99b');\n",
              "          const dataTable =\n",
              "            await google.colab.kernel.invokeFunction('convertToInteractive',\n",
              "                                                     [key], {});\n",
              "          if (!dataTable) return;\n",
              "\n",
              "          const docLinkHtml = 'Like what you see? Visit the ' +\n",
              "            '<a target=\"_blank\" href=https://colab.research.google.com/notebooks/data_table.ipynb>data table notebook</a>'\n",
              "            + ' to learn more about interactive tables.';\n",
              "          element.innerHTML = '';\n",
              "          dataTable['output_type'] = 'display_data';\n",
              "          await google.colab.output.renderOutput(dataTable, element);\n",
              "          const docLink = document.createElement('div');\n",
              "          docLink.innerHTML = docLinkHtml;\n",
              "          element.appendChild(docLink);\n",
              "        }\n",
              "      </script>\n",
              "    </div>\n",
              "  </div>\n",
              "  "
            ]
          },
          "metadata": {},
          "execution_count": 87
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "from sklearn.compose import make_column_transformer\n",
        "from sklearn.preprocessing import MinMaxScaler, OneHotEncoder\n",
        "\n",
        "# create an instance of column transformer \n",
        "ct = make_column_transformer(\n",
        "    (MinMaxScaler(), [\"age\",\"bmi\",\"children\"]), # get all values from 0 to 1\n",
        "    (OneHotEncoder(handle_unknown=\"ignore\"), [\"sex\", \"smoker\", \"region\"]))\n",
        "\n",
        "# Create X & y\n",
        "X = insurance.drop(\"charges\", axis = 1)\n",
        "y = insurance[\"charges\"]\n",
        "\n",
        "# Build our train and test sets\n",
        "X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.2)\n",
        "\n",
        "# Fit column transformer to the training data only (doing so on test data would result in data leakage)\n",
        "ct.fit(X_train)\n",
        "\n",
        "# Transform training and test data with normalization (MinMaxScalar) and one hot encoding (OneHotEncoder)\n",
        "X_train_normal = ct.transform(X_train)\n",
        "X_test_normal = ct.transform(X_test)"
      ],
      "metadata": {
        "id": "w0O40q1m-4Un"
      },
      "execution_count": 88,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "# let's see how our data look\n",
        "X_train_normal[0]"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "OoEHEVeQ_I89",
        "outputId": "65f3971b-e421-47bb-ccd4-12597f445df9"
      },
      "execution_count": 90,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "array([0.19565217, 0.39359699, 0.2       , 1.        , 0.        ,\n",
              "       1.        , 0.        , 1.        , 0.        , 0.        ,\n",
              "       0.        ])"
            ]
          },
          "metadata": {},
          "execution_count": 90
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "X_test_normal[0]"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "9u80pO60B-DU",
        "outputId": "e1cd3417-3c33-4b89-8462-363df3a850e4"
      },
      "execution_count": 91,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "array([0.17391304, 0.29862793, 0.        , 0.        , 1.        ,\n",
              "       0.        , 1.        , 0.        , 0.        , 1.        ,\n",
              "       0.        ])"
            ]
          },
          "metadata": {},
          "execution_count": 91
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "#it's ready to model now, let do it\n",
        "\n",
        "tf.random.set_seed(42)\n",
        "\n",
        "# create a model\n",
        "insurance_model_3 = tf.keras.Sequential([\n",
        "    tf.keras.layers.Dense(100),\n",
        "    tf.keras.layers.Dense(10),\n",
        "    tf.keras.layers.Dense(1)\n",
        "])\n",
        "\n",
        "# compile the model\n",
        "insurance_model_3.compile(loss=tf.keras.losses.mae,\n",
        "                          optimizer = tf.keras.optimizers.Adam(),\n",
        "                          metrics=[\"mae\"])\n",
        "\n",
        "# fit the model in normalized train data\n",
        "insurance_model_3.fit(X_train_normal, y_train, epochs = 200, verbose=0)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "3LgEF4TpCDhG",
        "outputId": "67924f19-a175-4d34-c61c-d34417fcccc1"
      },
      "execution_count": 92,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "<keras.callbacks.History at 0x7fe5b6d28550>"
            ]
          },
          "metadata": {},
          "execution_count": 92
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "# let's evaluate our model\n",
        "insurance_model_3.evaluate(X_test_normal, y_test)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "9CzVIEWSCy7c",
        "outputId": "76c39723-5697-42b0-8549-0679632ec435"
      },
      "execution_count": 93,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "9/9 [==============================] - 0s 2ms/step - loss: 3340.7031 - mae: 3340.7031\n"
          ]
        },
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "[3340.703125, 3340.703125]"
            ]
          },
          "metadata": {},
          "execution_count": 93
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 206
        },
        "id": "GmxJNFApC6lb",
        "outputId": "67e586b1-13f3-41d4-e97a-faa1bfd7cc1b"
      },
      "execution_count": 95,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "      age     sex    bmi  children smoker     region\n",
              "1274   26    male  27.06         0    yes  southeast\n",
              "1223   20  female  24.42         0    yes  southeast\n",
              "221    53  female  33.25         0     no  northeast\n",
              "1129   19  female  18.60         0     no  southwest\n",
              "794    39  female  31.92         2     no  northwest"
            ],
            "text/html": [
              "\n",
              "  <div id=\"df-ec5ee25e-7e60-4357-a143-18c9ed5cd771\">\n",
              "    <div class=\"colab-df-container\">\n",
              "      <div>\n",
              "<style scoped>\n",
              "    .dataframe tbody tr th:only-of-type {\n",
              "        vertical-align: middle;\n",
              "    }\n",
              "\n",
              "    .dataframe tbody tr th {\n",
              "        vertical-align: top;\n",
              "    }\n",
              "\n",
              "    .dataframe thead th {\n",
              "        text-align: right;\n",
              "    }\n",
              "</style>\n",
              "<table border=\"1\" class=\"dataframe\">\n",
              "  <thead>\n",
              "    <tr style=\"text-align: right;\">\n",
              "      <th></th>\n",
              "      <th>age</th>\n",
              "      <th>sex</th>\n",
              "      <th>bmi</th>\n",
              "      <th>children</th>\n",
              "      <th>smoker</th>\n",
              "      <th>region</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>1274</th>\n",
              "      <td>26</td>\n",
              "      <td>male</td>\n",
              "      <td>27.06</td>\n",
              "      <td>0</td>\n",
              "      <td>yes</td>\n",
              "      <td>southeast</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1223</th>\n",
              "      <td>20</td>\n",
              "      <td>female</td>\n",
              "      <td>24.42</td>\n",
              "      <td>0</td>\n",
              "      <td>yes</td>\n",
              "      <td>southeast</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>221</th>\n",
              "      <td>53</td>\n",
              "      <td>female</td>\n",
              "      <td>33.25</td>\n",
              "      <td>0</td>\n",
              "      <td>no</td>\n",
              "      <td>northeast</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1129</th>\n",
              "      <td>19</td>\n",
              "      <td>female</td>\n",
              "      <td>18.60</td>\n",
              "      <td>0</td>\n",
              "      <td>no</td>\n",
              "      <td>southwest</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>794</th>\n",
              "      <td>39</td>\n",
              "      <td>female</td>\n",
              "      <td>31.92</td>\n",
              "      <td>2</td>\n",
              "      <td>no</td>\n",
              "      <td>northwest</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>\n",
              "      <button class=\"colab-df-convert\" onclick=\"convertToInteractive('df-ec5ee25e-7e60-4357-a143-18c9ed5cd771')\"\n",
              "              title=\"Convert this dataframe to an interactive table.\"\n",
              "              style=\"display:none;\">\n",
              "        \n",
              "  <svg xmlns=\"http://www.w3.org/2000/svg\" height=\"24px\"viewBox=\"0 0 24 24\"\n",
              "       width=\"24px\">\n",
              "    <path d=\"M0 0h24v24H0V0z\" fill=\"none\"/>\n",
              "    <path d=\"M18.56 5.44l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94zm-11 1L8.5 8.5l.94-2.06 2.06-.94-2.06-.94L8.5 2.5l-.94 2.06-2.06.94zm10 10l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94z\"/><path d=\"M17.41 7.96l-1.37-1.37c-.4-.4-.92-.59-1.43-.59-.52 0-1.04.2-1.43.59L10.3 9.45l-7.72 7.72c-.78.78-.78 2.05 0 2.83L4 21.41c.39.39.9.59 1.41.59.51 0 1.02-.2 1.41-.59l7.78-7.78 2.81-2.81c.8-.78.8-2.07 0-2.86zM5.41 20L4 18.59l7.72-7.72 1.47 1.35L5.41 20z\"/>\n",
              "  </svg>\n",
              "      </button>\n",
              "      \n",
              "  <style>\n",
              "    .colab-df-container {\n",
              "      display:flex;\n",
              "      flex-wrap:wrap;\n",
              "      gap: 12px;\n",
              "    }\n",
              "\n",
              "    .colab-df-convert {\n",
              "      background-color: #E8F0FE;\n",
              "      border: none;\n",
              "      border-radius: 50%;\n",
              "      cursor: pointer;\n",
              "      display: none;\n",
              "      fill: #1967D2;\n",
              "      height: 32px;\n",
              "      padding: 0 0 0 0;\n",
              "      width: 32px;\n",
              "    }\n",
              "\n",
              "    .colab-df-convert:hover {\n",
              "      background-color: #E2EBFA;\n",
              "      box-shadow: 0px 1px 2px rgba(60, 64, 67, 0.3), 0px 1px 3px 1px rgba(60, 64, 67, 0.15);\n",
              "      fill: #174EA6;\n",
              "    }\n",
              "\n",
              "    [theme=dark] .colab-df-convert {\n",
              "      background-color: #3B4455;\n",
              "      fill: #D2E3FC;\n",
              "    }\n",
              "\n",
              "    [theme=dark] .colab-df-convert:hover {\n",
              "      background-color: #434B5C;\n",
              "      box-shadow: 0px 1px 3px 1px rgba(0, 0, 0, 0.15);\n",
              "      filter: drop-shadow(0px 1px 2px rgba(0, 0, 0, 0.3));\n",
              "      fill: #FFFFFF;\n",
              "    }\n",
              "  </style>\n",
              "\n",
              "      <script>\n",
              "        const buttonEl =\n",
              "          document.querySelector('#df-ec5ee25e-7e60-4357-a143-18c9ed5cd771 button.colab-df-convert');\n",
              "        buttonEl.style.display =\n",
              "          google.colab.kernel.accessAllowed ? 'block' : 'none';\n",
              "\n",
              "        async function convertToInteractive(key) {\n",
              "          const element = document.querySelector('#df-ec5ee25e-7e60-4357-a143-18c9ed5cd771');\n",
              "          const dataTable =\n",
              "            await google.colab.kernel.invokeFunction('convertToInteractive',\n",
              "                                                     [key], {});\n",
              "          if (!dataTable) return;\n",
              "\n",
              "          const docLinkHtml = 'Like what you see? Visit the ' +\n",
              "            '<a target=\"_blank\" href=https://colab.research.google.com/notebooks/data_table.ipynb>data table notebook</a>'\n",
              "            + ' to learn more about interactive tables.';\n",
              "          element.innerHTML = '';\n",
              "          dataTable['output_type'] = 'display_data';\n",
              "          await google.colab.output.renderOutput(dataTable, element);\n",
              "          const docLink = document.createElement('div');\n",
              "          docLink.innerHTML = docLinkHtml;\n",
              "          element.appendChild(docLink);\n",
              "        }\n",
              "      </script>\n",
              "    </div>\n",
              "  </div>\n",
              "  "
            ]
          },
          "metadata": {},
          "execution_count": 95
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "# Do some Exercises"
      ],
      "metadata": {
        "id": "eUzuZt2uDNc0"
      }
    },
    {
      "cell_type": "code",
      "source": [],
      "metadata": {
        "id": "siQAH3EqDkjl"
      },
      "execution_count": null,
      "outputs": []
    }
  ]
}