model.ipynb

{
  "nbformat": 4,
  "nbformat_minor": 0,
  "metadata": {
    "colab": {
      "name": "model.ipynb",
      "provenance": [],
      "collapsed_sections": [],
      "mount_file_id": "1sTEdurpWREwnkjrad-P8Nznol7JbxXfD",
      "authorship_tag": "ABX9TyMeR2hfxt04YikeELVkq0j0"
    },
    "kernelspec": {
      "name": "python3",
      "display_name": "Python 3"
    }
  },
  "cells": [
    {
      "cell_type": "code",
      "metadata": {
        "id": "JTLXWvEX2n0Y"
      },
      "source": [
        "# !pip install --upgrade jsonpickle tweepy"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "0hJb4m5O7V0y",
        "outputId": "487b2bc8-3397-40db-c628-1bcf47610e82",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 35
        }
      },
      "source": [
        "import os\n",
        "os.chdir('drive/My Drive/twitter-bot-detection')\n",
        "os.getcwd()"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "application/vnd.google.colaboratory.intrinsic+json": {
              "type": "string"
            },
            "text/plain": [
              "'/content/drive/My Drive/twitter-bot-detection'"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 6
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "WQSCCljYq6X2"
      },
      "source": [
        "import sys\n",
        "import jsonpickle\n",
        "import numpy as np\n",
        "import pandas as pd\n",
        "import sklearn\n",
        "from sklearn import preprocessing\n",
        "from sklearn.preprocessing import PolynomialFeatures,MinMaxScaler\n",
        "from sklearn.metrics import r2_score\n",
        "from sklearn.metrics import accuracy_score\n",
        "from sklearn.linear_model import LogisticRegression\n",
        "from sklearn.linear_model import LogisticRegressionCV\n",
        "from sklearn.pipeline import make_pipeline\n",
        "from sklearn.preprocessing import PolynomialFeatures\n",
        "from sklearn.model_selection import train_test_split\n",
        "from pandas.plotting import scatter_matrix\n",
        "from sklearn.discriminant_analysis import LinearDiscriminantAnalysis\n",
        "from sklearn.discriminant_analysis import QuadraticDiscriminantAnalysis\n",
        "from sklearn.ensemble import AdaBoostClassifier\n",
        "from sklearn.tree import DecisionTreeClassifier\n",
        "from sklearn.ensemble import RandomForestClassifier\n",
        "from sklearn.model_selection import train_test_split\n",
        "from sklearn.metrics import classification_report,confusion_matrix\n",
        "from sklearn.model_selection import GridSearchCV\n",
        "from sklearn.ensemble import GradientBoostingClassifier\n",
        "import xgboost as xgb\n",
        "from sklearn.metrics import confusion_matrix\n",
        "from sklearn.neighbors import KNeighborsClassifier\n",
        "from sklearn.model_selection import cross_val_score\n",
        "from sklearn.model_selection import KFold\n",
        "from sklearn.svm import SVC\n",
        "\n",
        "from sklearn import svm\n",
        "import seaborn as sns\n",
        "import matplotlib.pyplot as plt\n",
        "import matplotlib\n",
        "%matplotlib inline\n",
        "\n",
        "import warnings\n",
        "warnings.filterwarnings('ignore')"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "XbOFvwIT6sA3"
      },
      "source": [
        "bot_df = pd.read_csv(r\"data/final_bots.csv\",index_col='User ID')\n",
        "user_df = pd.read_csv(r\"data/final_users.csv\",index_col='User ID')"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "wTqs74VY7PhP"
      },
      "source": [
        "bot_df['bot']=1\n",
        "user_df['bot']=0\n",
        "total_df = bot_df.append(user_df)"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "OwRd7_OR7fHg"
      },
      "source": [
        "train_data, test_data = train_test_split(total_df, test_size = 0.2, random_state=87)\n",
        "\n",
        "Y_train=train_data['bot']\n",
        "Y_test=test_data['bot']\n",
        "X_train=train_data.drop('bot',axis=1)\n",
        "X_test=test_data.drop('bot',axis=1)\n",
        "X_train = X_train.drop(['Account age (days)'], axis=1)\n",
        "X_test = X_test.drop(['Account age (days)'], axis=1)"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "nZIuWJob7w0n",
        "outputId": "c6233fab-88a9-48bd-c267-652a4ff1c04f",
        "colab": {
          "base_uri": "https://localhost:8080/"
        }
      },
      "source": [
        "print(np.count_nonzero(Y_test)/len(Y_test))\n",
        "print(np.count_nonzero(Y_train)/len(Y_train))"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "0.45491643995336184\n",
            "0.45430251202565475\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "RdyKxx5g7_cy"
      },
      "source": [
        "# def normalize(df,df_train):\n",
        "#     result = df.copy()\n",
        "#     for feature_name in df_train.columns:\n",
        "#         max_value = df_train[feature_name].max()\n",
        "#         min_value = df_train[feature_name].min()\n",
        "#         result[feature_name] = (df[feature_name] - min_value) / (max_value - min_value)\n",
        "#     return result\n",
        "\n",
        "# X_train_scaled=normalize(X_train,X_train)\n",
        "# X_test_scaled=normalize(X_test,X_train)\n",
        "# X_train_scaled.describe()"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "0NvYc4qssGh-",
        "outputId": "995e8901-2fbe-4cce-b6ae-033640c78a11",
        "colab": {
          "base_uri": "https://localhost:8080/"
        }
      },
      "source": [
        "scaler = MinMaxScaler()\n",
        "scaler.fit(X_train)"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "MinMaxScaler(copy=True, feature_range=(0, 1))"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 13
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "TID0F_Z08N9V"
      },
      "source": [
        "X_train_scaled=scaler.transform(X_train)\n",
        "X_train_scaled = pd.DataFrame(X_train_scaled,columns=X_train.columns)\n",
        "X_test_scaled=scaler.transform(X_test)\n",
        "X_test_scaled = pd.DataFrame(X_test_scaled,columns=X_train.columns)"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "GLrg3VEbvG1Y",
        "outputId": "a61adfa1-222a-4cb6-8e60-48e9338fe63e",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 365
        }
      },
      "source": [
        "X_train_scaled.describe()"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/html": [
              "<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>Screen name length</th>\n",
              "      <th>Number of digits in screen name</th>\n",
              "      <th>User name length</th>\n",
              "      <th>Default profile (binary)</th>\n",
              "      <th>Default picture (binary)</th>\n",
              "      <th>Number of unique profile descriptions</th>\n",
              "      <th>Number of friends</th>\n",
              "      <th>Number of followers</th>\n",
              "      <th>Number of favorites</th>\n",
              "      <th>Number of tweets per hour</th>\n",
              "      <th>Number of tweets total</th>\n",
              "      <th>timing_tweet</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>count</th>\n",
              "      <td>20581.000000</td>\n",
              "      <td>20581.000000</td>\n",
              "      <td>20581.000000</td>\n",
              "      <td>20581.000000</td>\n",
              "      <td>20581.000000</td>\n",
              "      <td>20581.000000</td>\n",
              "      <td>20581.000000</td>\n",
              "      <td>20581.000000</td>\n",
              "      <td>20581.000000</td>\n",
              "      <td>20581.000000</td>\n",
              "      <td>20581.000000</td>\n",
              "      <td>20581.000000</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>mean</th>\n",
              "      <td>0.680138</td>\n",
              "      <td>0.026441</td>\n",
              "      <td>0.221217</td>\n",
              "      <td>0.128711</td>\n",
              "      <td>0.026627</td>\n",
              "      <td>0.047619</td>\n",
              "      <td>0.000891</td>\n",
              "      <td>0.000184</td>\n",
              "      <td>0.004234</td>\n",
              "      <td>0.008021</td>\n",
              "      <td>0.007801</td>\n",
              "      <td>0.717822</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>std</th>\n",
              "      <td>0.207034</td>\n",
              "      <td>0.078484</td>\n",
              "      <td>0.104966</td>\n",
              "      <td>0.334888</td>\n",
              "      <td>0.160993</td>\n",
              "      <td>0.091744</td>\n",
              "      <td>0.008337</td>\n",
              "      <td>0.007207</td>\n",
              "      <td>0.019174</td>\n",
              "      <td>0.022343</td>\n",
              "      <td>0.022043</td>\n",
              "      <td>0.178786</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>min</th>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>25%</th>\n",
              "      <td>0.538462</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.163265</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000059</td>\n",
              "      <td>0.000002</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000162</td>\n",
              "      <td>0.000152</td>\n",
              "      <td>0.588563</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>50%</th>\n",
              "      <td>0.692308</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.224490</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000173</td>\n",
              "      <td>0.000007</td>\n",
              "      <td>0.000030</td>\n",
              "      <td>0.001813</td>\n",
              "      <td>0.001737</td>\n",
              "      <td>0.733345</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>75%</th>\n",
              "      <td>0.846154</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.285714</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.055556</td>\n",
              "      <td>0.000526</td>\n",
              "      <td>0.000026</td>\n",
              "      <td>0.001232</td>\n",
              "      <td>0.007701</td>\n",
              "      <td>0.007168</td>\n",
              "      <td>0.869844</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>max</th>\n",
              "      <td>1.000000</td>\n",
              "      <td>1.000000</td>\n",
              "      <td>1.000000</td>\n",
              "      <td>1.000000</td>\n",
              "      <td>1.000000</td>\n",
              "      <td>1.000000</td>\n",
              "      <td>1.000000</td>\n",
              "      <td>1.000000</td>\n",
              "      <td>1.000000</td>\n",
              "      <td>1.000000</td>\n",
              "      <td>1.000000</td>\n",
              "      <td>1.000000</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>"
            ],
            "text/plain": [
              "       Screen name length  ...  timing_tweet\n",
              "count        20581.000000  ...  20581.000000\n",
              "mean             0.680138  ...      0.717822\n",
              "std              0.207034  ...      0.178786\n",
              "min              0.000000  ...      0.000000\n",
              "25%              0.538462  ...      0.588563\n",
              "50%              0.692308  ...      0.733345\n",
              "75%              0.846154  ...      0.869844\n",
              "max              1.000000  ...      1.000000\n",
              "\n",
              "[8 rows x 12 columns]"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 169
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "ZD-GK8p_8g5Z",
        "outputId": "1ef9a42a-fa42-4ca0-a4ff-e3b624e08423",
        "colab": {
          "base_uri": "https://localhost:8080/"
        }
      },
      "source": [
        "# Logistic Regression\n",
        "logreg = LogisticRegression(C=100000,fit_intercept=True).fit(X_train_scaled,Y_train)\n",
        "logreg_train = logreg.score(X_train_scaled, Y_train)\n",
        "\n",
        "print('Accuracy of logistic regression model on training set is {:.3f}'.format(logreg_train))\n",
        "# Classification error on test set\n",
        "lr_pred = logreg.predict(X_test_scaled)\n",
        "logreg_test = logreg.score(X_test_scaled, Y_test)\n",
        "print('Accuracy of logistic regression model on the test set is {:.3f}'.format(logreg_test))"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Accuracy of logistic regression model on training set is 0.765\n",
            "Accuracy of logistic regression model on the test set is 0.751\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "90RiNVQ5ahQb",
        "outputId": "e4783ec5-da82-43f8-87b3-3e60efe7d512",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 283
        }
      },
      "source": [
        "cf_matrix = confusion_matrix(lr_pred,Y_test)\n",
        "sns.heatmap(cf_matrix/np.sum(cf_matrix), annot=True, \n",
        "            fmt='.2%', cmap='Blues')"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "<matplotlib.axes._subplots.AxesSubplot at 0x7f9ce5e1b5c0>"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 163
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 432x288 with 2 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "dEmXBPcM80O5",
        "outputId": "99bfb5a3-27e0-4bb2-ece1-a5b62d0924e2",
        "colab": {
          "base_uri": "https://localhost:8080/"
        }
      },
      "source": [
        "# Logistic regression w/ quadratic + interaction terms + regularization\n",
        "polynomial_logreg_estimator = make_pipeline(\n",
        "    PolynomialFeatures(degree=2, include_bias=True),\n",
        "    LogisticRegressionCV(multi_class=\"ovr\", penalty='l2', cv=5, max_iter=10000))\n",
        "linearLogCVpoly = polynomial_logreg_estimator.fit(X_train_scaled, Y_train)\n",
        "# Compare results\n",
        "print('Polynomial-logistic accuracy: train={:.1%}, test={:.1%}'.format(\n",
        "    linearLogCVpoly.score(X_train_scaled, Y_train), linearLogCVpoly.score(X_test_scaled, Y_test)))\n",
        "linearLogCVpoly_train = linearLogCVpoly.score(X_train_scaled, Y_train)\n",
        "linearLogCVpoly_test = linearLogCVpoly.score(X_test_scaled, Y_test)"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Polynomial-logistic accuracy: train=78.3%, test=77.0%\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "lT5IFTEr-UB9",
        "outputId": "537d1068-ddf7-4a63-9bc8-079ccbab0b5b",
        "colab": {
          "base_uri": "https://localhost:8080/"
        }
      },
      "source": [
        "# LDA and QDA Model\n",
        "lda = LinearDiscriminantAnalysis(store_covariance=True)\n",
        "qda = QuadraticDiscriminantAnalysis(store_covariance=True)\n",
        "lda.fit(X_train_scaled, Y_train)\n",
        "qda.fit(X_train_scaled, Y_train)\n",
        "lda.predict(X_test_scaled)\n",
        "qda.predict(X_test_scaled)\n",
        "\n",
        "print('LDA accuracy train={:.1%}, test: {:.1%}'.format(\n",
        "    lda.score(X_train_scaled, Y_train), lda.score(X_test_scaled, Y_test)))\n",
        "\n",
        "lda_train = lda.score(X_train_scaled, Y_train)\n",
        "lda_test = lda.score(X_test_scaled, Y_test)\n",
        "\n",
        "print('QDA accuracy train={:.1%}, test: {:.1%}'.format(\n",
        "    qda.score(X_train_scaled, Y_train), qda.score(X_test_scaled, Y_test)))\n",
        "\n",
        "qda_train = qda.score(X_train_scaled, Y_train)\n",
        "qda_test = qda.score(X_test_scaled, Y_test)"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "LDA accuracy train=67.1%, test: 66.0%\n",
            "QDA accuracy train=70.2%, test: 69.2%\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "ircBJLl0-lT4",
        "outputId": "64e00846-4507-404f-e1d2-6ff27de5feb5",
        "colab": {
          "base_uri": "https://localhost:8080/"
        }
      },
      "source": [
        "# Random Forest\n",
        "ntrees = 50\n",
        "rf = RandomForestClassifier(n_estimators=ntrees , max_depth=15, max_features='auto')\n",
        "rf.fit(X_train_scaled, Y_train)\n",
        "rf_train =rf.score(X_train_scaled, Y_train)\n",
        "rf_test =rf.score(X_test_scaled, Y_test)\n",
        "\n",
        "print('RF accuracy train={:.1%}, test: {:.1%}'.format(rf_train,rf_test))"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "RF accuracy train=97.8%, test: 89.3%\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "yYgvjC-H-2ql",
        "outputId": "1b3f60bd-4bce-4b93-e68e-e157a923ba04",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 368
        }
      },
      "source": [
        "plt.figure(figsize=(5,5))\n",
        "plt.title('Variable Importance from Random Forest')\n",
        "plt.xlabel('Relative Importance')\n",
        "pd.Series(rf.feature_importances_,index=list(X_train_scaled)).sort_values().plot(kind=\"barh\")"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "<matplotlib.axes._subplots.AxesSubplot at 0x7f9ce61e1940>"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 93
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 360x360 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "fcdV7gYJfmgi",
        "outputId": "9700aef5-5317-4a13-895e-b70822abb257",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 287
        }
      },
      "source": [
        "rf_pred = rf.predict(X_test_scaled)\n",
        "cf_matrix = confusion_matrix(rf_pred,Y_test)\n",
        "sns.heatmap(cf_matrix/np.sum(cf_matrix), annot=True, \n",
        "            fmt='.2%', cmap='Blues')"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "<matplotlib.axes._subplots.AxesSubplot at 0x7f9ce5ec86d8>"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 121
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "iVBORw0KGgoAAAANSUhEUgAAAV0AAAD8CAYAAADUv3dIAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADh0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uMy4yLjIsIGh0dHA6Ly9tYXRwbG90bGliLm9yZy+WH4yJAAAY8klEQVR4nO3deXgV1f3H8ff33hAIsqNBNhUQwbD7YxFU5IdoQVmsWAuCVYtEUapFEFAr/qRVEFyqgktUXMpatNYgCFqURRENrRQbIIKoEDCyCAZZs5zfH4lpICE3gWQyDJ+XzzxmZs6cOcOT5/OcfGfmXnPOISIi3giV9wBERE4lCl0REQ8pdEVEPKTQFRHxkEJXRMRDCl0REQ8pdEVEjsHMeppZipltNLOxhey/ycx2mNnq3OWWSH1Glc1QRURObmYWBqYClwOpQJKZJTrn1h7VdI5zbnhx+9VMV0SkcB2Bjc65Tc65w8BsoN+JdlrmM92YdsP1ypsUkLbi6fIegvhQ9ZiQnWgfJcmcg6un3grE59uU4JxLyP25PrAl375UoFMh3fQ3s67Al8AI59yWQtrkUXlBRE5ZuQGbELHhsc0DZjnnDpnZrcBrQPeiDlB5QUSCxULFX4q2FWiYb71B7rY8zrldzrlDuasvAf8TqVOFrogESyhc/KVoSUBTM2tkZtHAACAxfwMzq5tvtS+wLlKnKi+ISLDYCZeFAXDOZZrZcGAREAamOeeSzWw8sMo5lwjcaWZ9gUzgB+CmSP0qdEUkWCKXDYrNObcAWHDUtnH5fr4XuLckfSp0RSRYSmmmW1YUuiISLKU40y0LCl0RCRbNdEVEPBT5qYRypdAVkWBReUFExEMqL4iIeEgzXRERDyl0RUQ8FNaNNBER76imKyLiIZUXREQ8pJmuiIiHNNMVEfGQZroiIh7Sa8AiIh5SeUFExEMqL4iIeEgzXRERDyl0RUQ8pBtpIiIeUk1XRMRDKi+IiHhIM10REe+YQldExDsKXRERD1lIoSsi4hnNdEVEPKTQFRHxkEJXRMRL/s5cha6IBItmuiIiHgqF9EaaiIhnNNMVEfGSvzNXoSsiwaKZroiIhxS6IiIe0mvAIiIe0kxXRMRDfg9dfz/QJiJSQmZW7KUYffU0sxQz22hmY4to19/MnJm1j9SnZroiEiilNdM1szAwFbgcSAWSzCzRObf2qHZVgbuAT4vTr2a6IhIsVoKlaB2Bjc65Tc65w8BsoF8h7f4IPAocLM7wFLoiEiihUKjYi5nFm9mqfEt8vq7qA1vyrafmbstjZhcADZ1z84s7PpUXRCRQSlJecM4lAAnHeZ4Q8ARwU0mOU+gW0/r5D7F33yGysrPJzMrm4kGTqFmtMn959LecXa8W3277gcGjX2bP3gMFjh3UpxNjb/kFABNfWsSMeZ8SU6kCMyYNoXGD08nKdixY9gUPPJ0IwLABlzKk/0VsSdvNdSMSyMjMokvbxlx9WVtGP/43T69bim9vejoPj3+ArzZuwMz4w//9idZt2uXtXzh/Hq+/+hLOOSpXPo0x9z/Iec2a5+3Pysrixut/xRmxsTz5zPMAPHDvPXy18UsuvqQbt985AoCXX3yOJk2a0q17D28v8GRReg8vbAUa5ltvkLvtZ1WBlsCS3KA/E0g0s77OuVXH6lTlhRLoGf8UFw6YyMWDJgEw6ubLWfJZCq36jWfJZymMuvmKAsfUrFaZ++N70fWGx7hk8GTuj+9FjaoxAPz59cW0veZPXDhgIp3bNOaKi+IAGNCrPR2um8DKf2/i8i7nAzB2aC8mvLjQoyuV4/H4pEe4sMvFzP37Amb89S0aNWpyxP569Rvw/MuvM+uNRIbED2PCHx88Yv/smX/hnEaN89Y3fJlCxUoVmTn3bdYmf8FPe/eyc8d2kr9Yo8AtQik+vZAENDWzRmYWDQwAEn/e6Zz70Tl3unPuHOfcOcBKoMjABYXuCendrTXT5+XcsJw+71P6/G/rAm0u73I+i1euZ3f6fvbsPcDileu54qI4DhzMYNmqDQBkZGaxev0W6sfWAHJ+aSpEhalcKZqMzCwGXtWB9z5OZnf6fu8uTkrkp717+fxfq+j3y2sBqFAhmqrVqh3RpnXbdlSrVh2Alq3bsP37tLx933+fxsfLl9LvmmvztkVFRXHo4CGys7PJzMwkFA7xwrPPED9suAdXdPIqrdB1zmUCw4FFwDrgr865ZDMbb2Z9j3d8EcsLZtacnDt2PxeQtwKJzrl1x3vSk5FzjnnPDsc5x8tvfsy0v31MbO2qpO1MByBtZzqxtasWOK7eGTVI/X533vrW7Xuod0aNI9pUrxLDlV1bMWXmEgCem7OUpa+PZN1X3/HJ6k3MfTKePndMLbNrkxO3bWsqNWvWYvy4+9jwZQrN4+IYOfo+YmIqF9o+8a036XzxJXnrT06ewO9+P4r9+/blbWvUuAk1a9bkhgH9ubJ3X1I3bybbZdP8/BZlfj0ns9J8OcI5twBYcNS2ccdo2604fRYZumY2BhhIzqMSn+VubgDMMrPZzrmJxTlJEFx285Ns2/EjZ9SswjvPDyflm7QCbZwreb/hcIjXJt7Es7OW8M3WXQDMmp/ErPlJANwb35NnZy3lFxe1YFDvjqSm7WbME2/hjudkUmYys7JIWb+WUWPvp2WrNjz+6CO8Nu1FbrvjrgJtVyV9SuLf3yThlekALF/2ITVr1uL8uBb8M+mzI9rePfq+//585zDu/cNDTHvxeTZ8mUKnCztzdf/ryvbCTkJ+/+yFSOWFIUAH59xE59z03GUiOc+vDTnWQfkfw8jcmVya4y0323b8CMCO3T+R+MEaOrQ4h+279nLm6Tl/Qp55ejV2/LC3kOP20KBOzbz1+rE12LZjT9761D8M5KvNO/JmufnVPaM67Vucw7wla7jrhu4MHjONPXsP8L8dm5XuxckJi61Th9jYOrRs1QaA7pdfQcq6tQXabfgyhYcfeoDJf55CjRo5vxdrVn/O8qUf0q/XZdw/diSrkj5l3H2jjzhu6YeLaX5+C/Yf2MfW1C1MmPwki//xHgcPFLxxe6orzTfSykKk0M0G6hWyvW7uvkI55xKcc+2dc+2jTj/5/xSqXCmaKpUr5v3co3Nzkr/axvylXzC4TycABvfpxDtL1hQ49v0V6+jRuTk1qsZQo2oMPTo35/0VOZWZB2/vTfWqMYya/Gah5x13+1X88bl3AIipWAHnINs5KsdUKIvLlBNw+ulnEHtmXb795msAkj5dSaPG5x7RJu27bYwZeScP/elRzj67Ud72O+68m3feW8Lb7y7m4YmP075DJ8Y/Milvf2ZGBrNnvM5vbhrCoYOH8u7OZ2dnkZGRUfYXd5Lxe+hGqun+HlhsZhv470PCZwHnklNgPiXE1q7KnCeGAhAVDjPn3VW8v2Id/0zezPRHf8uNV3dm83c/MHj0NAAuiDuLW669mNvHz2R3+n4mvLiQj6bnzFweSVjI7vT91I+twdihPVm/KY1PZo0B4Pk5S3n1rU8AaNOsAQCr16cCMOfdVayaex+pabt54tV/eHr9Ujz3jLmfB+67h8yMDOrVb8i48Q/z5tzZAPT/1QBeSniWH/fs4dFHxgMQjgrz+sw3IvY7d85MrupzNZViYmh6XjMOHTzIwGv70uXirgVu1gn4/PNusEi1wdwHgDty5I20JOdcVnFOENNuuIqPUkDaiqfLewjiQ9VjTrwg2/SehcXOnA2Te3oe0RGfXnDOZZPz/JmIiO+FfH4jTW+kiUig+L28oNAVkUDRTFdExEOa6YqIeMjvX9ej0BWRQPF55ip0RSRYQiF/f46XQldEAkUzXRERD6mmKyLiIZ9nrkJXRIJFM10REQ/5PHMVuiISLHojTUTEQyoviIh4yOeZq9AVkWDRTFdExEM+z1yFrogEi26kiYh4SOUFEREPKXRFRDzk88xV6IpIsGimKyLiIZ9nrkJXRIJFTy+IiHgo5POprkJXRALF55mr0BWRYNGNNBERD/m8pKvQFZFg0Y00EREPGQpdERHP+Hyiq9AVkWDRjTQREQ/5PHMVuiISLH5/OSJU3gMQESlNoZAVe4nEzHqaWYqZbTSzsYXsv83MvjCz1Wb2kZnFRRzfcV6XiIgvmRV/KbofCwNTgV5AHDCwkFCd6Zxr5ZxrC0wCnog0PpUXRCRQSrG80BHY6JzbBGBms4F+wNqfGzjn0vO1Pw1wkTpV6IpIoJQkcs0sHojPtynBOZeQ+3N9YEu+falAp0L6uAO4G4gGukc6p0JXRAKlJI+M5QZsQsSGRfcxFZhqZtcDfwBuLKq9QldEAqUUX47YCjTMt94gd9uxzAaei9SpbqSJSKCU4tMLSUBTM2tkZtHAACAxfwMza5pv9SpgQ6RONdMVkUAprTfSnHOZZjYcWASEgWnOuWQzGw+scs4lAsPNrAeQAewmQmkBFLoiEjCl+dkLzrkFwIKjto3L9/NdJe1ToSsigaLPXhAR8ZC/I1ehKyIBE/b5ZzsqdEUkUFReEBHxkM8zV6ErIsHi9492VOiKSKD4PHPLPnR3J00p61PISajZ3fPKewjiQ98+3eeE+1BNV0TEQ2GFroiId3z+xJhCV0SCRaErIuIh1XRFRDykma6IiId8PtFV6IpIsET5PHUVuiISKD7PXIWuiASLXgMWEfGQzzNXoSsiwaKnF0REPKQPMRcR8ZDPM1ehKyLBYj7/ljSFrogEima6IiIeUuiKiHhIH3gjIuKhcKi8R1A0ha6IBIreSBMR8ZBquiIiHvL5RFehKyLBEtJzuiIi3tFMV0TEQ1E+L+oqdEUkUDTTFRHxkB4ZExHxkM8zV6ErIsHi8xfSFLoiEiwqL4iIeMjvoev3mbiISIlYCZaIfZn1NLMUM9toZmML2X+3ma01szVmttjMzo7Up0JXRALFrPhL0f1YGJgK9ALigIFmFndUs8+B9s651sAbwKRI41PoikigmFmxlwg6Ahudc5ucc4eB2UC//A2ccx865/bnrq4EGkTqVKErIoESKsESQX1gS7711NxtxzIEeDdSp7qRJiKBUpIbaWYWD8Tn25TgnEso6TnNbDDQHrg0UluFrogESkm+ric3YI8VsluBhvnWG+RuO/p8PYD7gUudc4cinVPlBREJlFIsLyQBTc2skZlFAwOAxPwNzKwd8ALQ1zm3vTjj00xXRAKltL6Y0jmXaWbDgUVAGJjmnEs2s/HAKudcIjAZqALMzT3vZudc36L6VeiKSKCU5qsRzrkFwIKjto3L93OPkvap0BWRQAn7/I00ha6IBIrPM1ehKyLBYvqONBER72imKyLiIX0bsIiIhzTTFRHxkN8/T1ehKyKB4vNvYFfoikiw6OkFEREP+by6oA+8OV5ZWVlc1/9qht9+6zHb/OO9RbRp0Yzk/3wBQMbhwzxw/730v7oPv/plX5I++xSAw4cPMyx+CNf0682cWTPyjh//4AOsW5tcthcix61iVIi3R17Mu2O68v693RjR6zwAujStzfx7uvLe2Et5fFBbwoX8vRtXvxpvjbiI9+/txsIxl9K7Xb28fY8NastHD17GgtFdWTC6K3H1qwHQq01d3r+3G3Pv6kKNyhUAOOv0yky56QIPrvbkYSX4rzxopnucZvzldRo3bsJP+34qdP++fT8xY/rrtGrdJm/bm2/Mzfn/3+exa9cu7rhtKDPnvMGKj5bT7oL/4Zb427hx8EB+PXAQKevXk5WdxflxLTy5Him5Q5nZDHzmE/YfziIqZLzx+4tYtn4Hjw9ux/VTPuHrHfu4+8pmXNuxAXNWbjni2AOHsxgxfTXf7NhHbLWKzL+nK8vWbyf9QCYAj7y9lgWrvzvimBu7nkOfx5bTq82ZXN2+Pq8u+4ZRVzXnsfkpnl3zycDvNV3NdI/D92lpLF+2hF/2v/aYbaY+/RQ3DxlKxYoV87Zt+mojHTt1AqB27dpUrVqV5P/8h6gKURw8eJDMzEyccznHP/Nn7vjdXWV7IXLC9h/OAiAqHKJCOERWtiMjK5uvd+wDYPn6HfRqU7fAcV/v2Mc3uW22px9i50+HqFWlYoF2+TkH0VEhKkWHychydGhcix3ph/L6kRwhs2Iv5TK+cjnrSW7SxEcYMfIeQqHC//nWrU0mLS2Nrpd2O2L7ec2as/TDD8jMzCQ1dQvr1ibzfdp3XNj5IrZt3crggddx/aAbWPLBYs6Pa0FsbB0PrkZORMhgweiu/OuRK1iesoPV3+4hHDJaNawOwJVt61K3ZkyRfbQ5qwbR4RDf7vxveI66qjkLx1zKA79sQXRUzu/Z1Pc3MHP4hfRoWYfEf27lzp7n8fSiL8vu4k5SpfltwGXhuMsLZnazc+6VY+zL+wqMKc++wJCh8YU1OyktXfIhtWrVIq5Fy7yabH7Z2dk8Nmki4x+eUGDf1df05+tNX3H9df2pW68ebdq2IxQOExUVxcTJjwOQkZHBsPghPDXlWSY/OoG0776jT99+dOt+WZlfm5RctoMrJy2jWkwUCbd04Ly6Vfndq/9i3DU5Ybl8/Q6yst0xj4+tVpEnb2jHyBmfk/tHDpPmrWN7+iGio0JM+HVrbuvRhKcXbuCjlJ30nrwcgGs6NODDtd/TOPY04rs34cf9Gfzfm8kczMjy4rJ9ze/P6drPf86W+ECzzc65syK1O5jJ8Z3Ap5568nHemfc2UeEoDh06xL59P9G9x+VMePQxAPbu3Uvvnj2oXPk0AHbu3EH16tV5aspztGjZ6oi+fjNoAA8+9CeanHtu3rYZf3mNKlWqElunDv9e/TlDbx3G0N/eyLTXpnt3kR5odve88h5CqbuzZ1MOHs4i4YNNedsuaX4GAzqfxR2v/LNA+yqVopjzuy5MfX9Dgfrtzy48tzbx3Zvw24TP8rZVqhDmlVs7csOzK5l2a0dufXkVV7atS4VwiNmfbC79C/PQt0/3OeHEXLlxT7Ez58Jza3ie0EXOdM1szbF2Aafk3753jRjJXSNGApD02ae89uq0vMAFqFq1Kks//u8MeMhNN3D3qNG0aNmKAwcO4JyjcuXKfLLiY8Lh8BGBm/7jjyxbuoTnEl5m6ZIP8r4m+uDBg95doBRbrSrRZGZlk34gk4oVQlzS7Aye+8dGaleJZtdPh4mOCjHssiZMeW9DgWMrhI2EIe15M2lLgcCNrVaR7ek5X7V1ReszSfku/Yj9t17WhFeWfk1mtqNShTDO5cy4Y6LDZXexJxN/T3QjlhfqAL8Adh+13YAVZTKik9TUZ56iRYuWRZYBfvhhF8PihxAKhYiNrcPDEycdsf+F56ZyS/xthEIhulx0CbNnzcx5vOzXA8p6+HIcYqtV5InB7XJvysA7q7fxQfJ27ut3Ppe1qIOZMf3jb1ixYRcArRpWZ/DFZzNm1hp6t6tHx3NrU+O0aK7tmPPdh6NmrGbt1nSe+s0F1KoSjQFrt6Zz35w1R5yz7dk1eGphTi331WVfM2/UJaQfyGDoS0me/xv40UldXjCzl4FXnHMfFbJvpnPu+kgnCFp5QUpHEMsLcuJKo7yQtOnHYmdOh8bV/VVecM4NKWJfxMAVEfGcvye6ejlCRIJFn70gIuIhn5d0FboiEiw+z1yFrogEi/l8qqvQFZFA8XnmKnRFJFh8nrkKXREJGJ+nrkJXRAJFj4yJiHhINV0REQ8pdEVEPKTygoiIhzTTFRHxkM8zV6ErIgHj89RV6IpIoPj9Q8wVuiISKP6OXIWuiASNz1NXoSsigaJHxkREPOTzki6h8h6AiEhpshIsEfsy62lmKWa20czGFrK/q5n9y8wyzeza4oxPoSsigWJmxV4i9BMGpgK9gDhgoJnFHdVsM3ATMLO441N5QUQCpRTLCx2Bjc65TTn92mygH7D25wbOuW9y92UXt1PNdEUkUEpSXjCzeDNblW+Jz9dVfWBLvvXU3G0nRDNdEQmWEsx0nXMJQEKZjaUQCl0RCZRSfGRsK9Aw33qD3G0nROUFEQkUs+IvESQBTc2skZlFAwOAxBMdn0JXRAIlZMVfiuKcywSGA4uAdcBfnXPJZjbezPoCmFkHM0sFfgW8YGbJkcan8oKIBEzpPb7gnFsALDhq27h8PyeRU3YoNoWuiASK399IU+iKSKD4PHMVuiISLJrpioh4KNLrveVNoSsigeLvyFXoikjA+Hyiq9AVkWDRh5iLiHjJ35mr0BWRYPF55ip0RSRY9BXsIiIe8nnm6gNvRES8pJmuiASK32e6Cl0RCRQ9MiYi4iHNdEVEPKTQFRHxkMoLIiIe0kxXRMRDPs9cha6IBIzPU1ehKyKB4vfXgM05V95jOGWYWbxzLqG8xyH+ot+LU4teA/ZWfHkPQHxJvxenEIWuiIiHFLoiIh5S6HpLdTspjH4vTiG6kSYi4iHNdEVEPKTQFRHxkELXI2bW08xSzGyjmY0t7/FI+TOzaWa23cz+U95jEe8odD1gZmFgKtALiAMGmllc+Y5KfOBVoGd5D0K8pdD1Rkdgo3Nuk3PuMDAb6FfOY5Jy5pxbBvxQ3uMQbyl0vVEf2JJvPTV3m4icYhS6IiIeUuh6YyvQMN96g9xtInKKUeh6IwloamaNzCwaGAAklvOYRKQcKHQ94JzLBIYDi4B1wF+dc8nlOyopb2Y2C/gEaGZmqWY2pLzHJGVPrwGLiHhIM10REQ8pdEVEPKTQFRHxkEJXRMRDCl0REQ8pdEVEPKTQFRHx0P8D9oOhwABag0EAAAAASUVORK5CYII=\n",
            "text/plain": [
              "<Figure size 432x288 with 2 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "Z9RtgdJmAVh8"
      },
      "source": [
        "# svm\n",
        "def fit_and_plot_svm_for_c(x_train, y_train, x_test, y_test, C):\n",
        "    # Fit SVM model\n",
        "    model = svm.SVC(C=C, kernel='linear')\n",
        "    model.fit(x_train, y_train)\n",
        "    \n",
        "    # Train and test error\n",
        "    tr_acc = model.score(x_train, y_train)\n",
        "    ts_acc = model.score(x_test, y_test)\n",
        "    return tr_acc, ts_acc"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "4Arw8ZfvAcXl"
      },
      "source": [
        "# Fit and plot for different 'C' values\n",
        "tr_acc, ts_acc = fit_and_plot_svm_for_c(X_train_scaled, Y_train,X_test_scaled, Y_test, 0.1)"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "wDA9K4S4Av3t",
        "outputId": "4f4e969a-5c57-4145-a3a0-63860f6a63ed",
        "colab": {
          "base_uri": "https://localhost:8080/"
        }
      },
      "source": [
        "print(tr_acc, ts_acc)"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "0.6387444730576746 0.6319471434123591\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "7-a6UqfmBJcd"
      },
      "source": [
        "# KNN\n",
        "from sklearn.neighbors import KNeighborsClassifier\n",
        "from sklearn.model_selection import cross_val_score\n",
        "from sklearn.model_selection import KFold\n",
        "\n",
        "scores_mean=[]\n",
        "scores_std=[]\n",
        "\n",
        "k_number=np.arange(1,40)\n",
        "\n",
        "for k in range(1,40):\n",
        "    knn = KNeighborsClassifier(n_neighbors = k)\n",
        "    score_mean= cross_val_score(knn,X_train_scaled,Y_train,cv=5).mean()\n",
        "    score_std=cross_val_score(knn,X_train_scaled,Y_train,cv=5).std()\n",
        "    scores_mean.append(score_mean)"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "cNn4UCjzBPDX",
        "outputId": "8595a60c-fbc2-4068-a044-d8bb708ef33e",
        "colab": {
          "base_uri": "https://localhost:8080/"
        }
      },
      "source": [
        "max_score_k=max(scores_mean)\n",
        "best_k=scores_mean.index(max(scores_mean))+1\n",
        "print('Best K=',best_k, 'with a max CV score of',max_score_k)\n",
        "\n",
        "knn_best_k = KNeighborsClassifier(n_neighbors = best_k)\n",
        "knn_best_k.fit(X_train_scaled,Y_train);\n",
        "\n",
        "pred_best_k = knn_best_k.predict(X_test_scaled)\n",
        "\n",
        "print('test accuracy',accuracy_score(Y_test, pred_best_k))\n",
        "\n",
        "knn_best_k_train = knn_best_k.score(X_train_scaled, Y_train)\n",
        "knn_best_k_test = knn_best_k.score(X_test_scaled, Y_test)"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Best K= 34 with a max CV score of 0.6800929469952387\n",
            "test accuracy 0.6706179556937427\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "MTRVopMZBTXY"
      },
      "source": [
        "# adaBoost\n",
        "adaboost = AdaBoostClassifier(base_estimator=DecisionTreeClassifier(max_depth=5), n_estimators=800, learning_rate=0.01)\n",
        "adaboost.fit(X_train_scaled, Y_train);"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "ZA0lV8DfOepD",
        "outputId": "086387c4-a926-4a1d-cf77-174c0bdeaacd",
        "colab": {
          "base_uri": "https://localhost:8080/"
        }
      },
      "source": [
        "y_pred_test = adaboost.predict(X_test_scaled)\n",
        "pred_adaboost = [round(value) for value in y_pred_test]\n",
        "adaboost_accuracy = accuracy_score(Y_test, pred_adaboost)\n",
        "print('test acc:',adaboost_accuracy)"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "test acc: 0.8907889623008162\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "M9XEaRExOhu5",
        "outputId": "6b51b56a-3bcc-4df7-8db7-144151e1fb3a",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 368
        }
      },
      "source": [
        "\n",
        "plt.figure(figsize=(5,5))\n",
        "plt.title('Variable Importance from Adaboosting')\n",
        "plt.xlabel('Relative Importance')\n",
        "pd.Series(adaboost.feature_importances_,index=list(X_train_scaled)).sort_values().plot(kind=\"barh\")"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "<matplotlib.axes._subplots.AxesSubplot at 0x7f9ce5f8c7b8>"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 103
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 360x360 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "eHBuw-v7OoP2",
        "outputId": "8d877434-51c8-4222-9164-28129fc777e2",
        "colab": {
          "base_uri": "https://localhost:8080/"
        }
      },
      "source": [
        "\n",
        "import xgboost as xgb\n",
        "from sklearn.metrics import confusion_matrix\n",
        "\n",
        "\n",
        "xgb = xgb.XGBClassifier(max_depth=5, n_estimators=300, learning_rate=0.01).fit(X_train_scaled, Y_train)\n",
        "y_pred_test_xgb = xgb.predict(X_test_scaled)\n",
        "accuracy_xgb = accuracy_score(Y_test, y_pred_test_xgb)\n",
        "\n",
        "\n",
        "print(\"test Accuracy: %.2f%%\" % (accuracy_xgb * 100.0))"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "test Accuracy: 89.35%\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "Pab5Q2LCOyoj",
        "outputId": "de2d81f1-df8f-46ad-c32d-d98a4b2b6c8a",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 368
        }
      },
      "source": [
        "plt.figure(figsize=(5,5))\n",
        "plt.title('Variable Importance from XGBoost')\n",
        "plt.xlabel('Relative Importance')\n",
        "pd.Series(xgb.feature_importances_,index=list(X_train_scaled)).sort_values().plot(kind=\"barh\")"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "<matplotlib.axes._subplots.AxesSubplot at 0x7f9ce5ed9b70>"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 105
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 360x360 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "nZhcN6tZR_-C"
      },
      "source": [
        "#Model Comparison"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "hm-STGx2QHk5",
        "outputId": "61c0508c-b3f2-42ce-9595-e94ec7b5216a",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 312
        }
      },
      "source": [
        "dict_var = {\n",
        "            \"test accuracy\": [logreg_test, linearLogCVpoly_test, rf_test, knn_best_k_test ,lda_test,qda_test, adaboost_accuracy, accuracy_xgb],\n",
        "            \"training accuracy\": [logreg_train, linearLogCVpoly_train, rf_train, knn_best_k_train ,lda_train,qda_train, adaboost_accuracy, accuracy_xgb],\n",
        "           }\n",
        "print(\"Performance comparison of the six methods:\")\n",
        "df_var = pd.DataFrame.from_dict(dict_var)\n",
        "df_var.index= ['Linear LR', 'Polynomial LR', 'Random Forest', 'kNN', 'LDA', 'QDA', 'AdaBoost', 'XGBoost']\n",
        "display(df_var)"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Performance comparison of the six methods:\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "display_data",
          "data": {
            "text/html": [
              "<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>test accuracy</th>\n",
              "      <th>training accuracy</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>Linear LR</th>\n",
              "      <td>0.750874</td>\n",
              "      <td>0.764734</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>Polynomial LR</th>\n",
              "      <td>0.770113</td>\n",
              "      <td>0.782566</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>Random Forest</th>\n",
              "      <td>0.892732</td>\n",
              "      <td>0.976775</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>kNN</th>\n",
              "      <td>0.670618</td>\n",
              "      <td>0.705748</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>LDA</th>\n",
              "      <td>0.659930</td>\n",
              "      <td>0.671202</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>QDA</th>\n",
              "      <td>0.691994</td>\n",
              "      <td>0.702104</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>AdaBoost</th>\n",
              "      <td>0.890789</td>\n",
              "      <td>0.890789</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>XGBoost</th>\n",
              "      <td>0.893510</td>\n",
              "      <td>0.893510</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>"
            ],
            "text/plain": [
              "               test accuracy  training accuracy\n",
              "Linear LR           0.750874           0.764734\n",
              "Polynomial LR       0.770113           0.782566\n",
              "Random Forest       0.892732           0.976775\n",
              "kNN                 0.670618           0.705748\n",
              "LDA                 0.659930           0.671202\n",
              "QDA                 0.691994           0.702104\n",
              "AdaBoost            0.890789           0.890789\n",
              "XGBoost             0.893510           0.893510"
            ]
          },
          "metadata": {
            "tags": []
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "QnIzztx0RsRk",
        "outputId": "79e1e7dc-3550-41cd-eaac-718c7af5a1b8",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 614
        }
      },
      "source": [
        "plt.figure(figsize=(14,8))\n",
        "xx = range(len(df_var))\n",
        "index_name=df_var.index\n",
        "plt.bar(xx, df_var['test accuracy'], color=(0.2, 0.4, 0.6, 0.6), alpha=0.5)\n",
        "plt.ylim(0.6,1)\n",
        "plt.title('Model Comparison on Test Set', fontsize=18)\n",
        "plt.ylabel('Test Accuracy (%)', fontsize=16)\n",
        "plt.xticks(xx,index_name,rotation=90,fontsize = 16);\n",
        "sns.despine()"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 1008x576 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "wXXNOc1-k1e7"
      },
      "source": [
        "#  export models to files\n",
        "import pickle\n",
        "models = [xgb,adaboost,rf,scaler]\n",
        "model_names = ['xgb','adaboost','rf','scaler']\n",
        "for i in range(len(models)):\n",
        "    pickle.dump(models[i], open('models/'+model_names[i], 'wb'))\n",
        " \n",
        " "
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "uO_d7COfwogQ"
      },
      "source": [
        "# import pickle\n",
        "# model_names = ['xgb','adaboost','rf','scaler']\n",
        "# for i in range(4):\n",
        "#     pickle.load(open('models/'+model_names[i], 'rb'))"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "gxhNSEGDTneb"
      },
      "source": [
        "#Neural Network"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "j67ZOSKnTmgP"
      },
      "source": [
        "\n",
        "import tensorflow.keras \n",
        "from tensorflow.keras import models\n",
        "from tensorflow.keras import layers\n",
        "\n",
        "\n",
        "from tensorflow.keras.models import Sequential\n",
        "from tensorflow.keras.layers import Dense\n",
        "from tensorflow.keras import regularizers\n",
        "\n",
        "from sklearn.metrics import r2_score as r2\n",
        "\n",
        "from sklearn.utils import shuffle\n",
        "from tensorflow.keras.layers import Dropout\n",
        "from tensorflow.keras.wrappers.scikit_learn import KerasClassifier\n",
        "\n",
        "# import tensorflow.contrib.eager as tfe\n",
        "# from sklearn.grid_search import GridSearchCV\n",
        "from sklearn.datasets import make_classification\n",
        "from sklearn.model_selection import cross_val_score"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "FzQGH8FFUPd7",
        "outputId": "00da6620-1444-4854-96ab-bbf42b7f15fe",
        "colab": {
          "base_uri": "https://localhost:8080/"
        }
      },
      "source": [
        "model_NN = models.Sequential()\n",
        "\n",
        "model_NN.add(layers.Dense(400, input_shape=(X_train_scaled.shape[1],),\n",
        "                activation='relu'))\n",
        "\n",
        "model_NN.add(layers.Dense(100, input_shape=(X_train_scaled.shape[1],),\n",
        "                activation='relu', \n",
        "                kernel_regularizer=regularizers.l2(0.01)))\n",
        "model_NN.add(Dropout(0.5))\n",
        "          \n",
        "\n",
        "\n",
        "model_NN.add(layers.Dense(1,  \n",
        "                activation='sigmoid')) \n",
        "\n",
        "model_NN.summary()"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Model: \"sequential_3\"\n",
            "_________________________________________________________________\n",
            "Layer (type)                 Output Shape              Param #   \n",
            "=================================================================\n",
            "dense_9 (Dense)              (None, 400)               5200      \n",
            "_________________________________________________________________\n",
            "dense_10 (Dense)             (None, 100)               40100     \n",
            "_________________________________________________________________\n",
            "dropout_3 (Dropout)          (None, 100)               0         \n",
            "_________________________________________________________________\n",
            "dense_11 (Dense)             (None, 1)                 101       \n",
            "=================================================================\n",
            "Total params: 45,401\n",
            "Trainable params: 45,401\n",
            "Non-trainable params: 0\n",
            "_________________________________________________________________\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "kUQAIL18Ujvm"
      },
      "source": [
        "tensorflow.keras.optimizers.Adam(lr=0.01, beta_1=0.9, beta_2=0.999, epsilon=None, decay=0.0, amsgrad=False)\n",
        "model_NN.compile(loss='binary_crossentropy', optimizer='Adam', metrics=['accuracy'])"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "8bOEOo1cU0dr"
      },
      "source": [
        "%%capture\n",
        "ANN_history = model_NN.fit(X_train_scaled, Y_train,batch_size=128,epochs=300,validation_split=0.2)"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "rvfdFKTfVVJ8",
        "outputId": "2ea0af33-432f-48a8-d009-010943b25edf",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 281
        }
      },
      "source": [
        "training_loss = ANN_history.history['loss']\n",
        "epoch_count = range(1, len(training_loss) + 1)\n",
        "plt.plot(epoch_count, training_loss, 'r--')\n",
        "plt.legend(['Training Loss'])\n",
        "plt.xlabel('Epoch')\n",
        "plt.ylabel('Loss')\n",
        "plt.show();"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "vumQApRnWeUQ",
        "outputId": "fc4f882f-84cb-4a8f-d008-3c44ef586f0b",
        "colab": {
          "base_uri": "https://localhost:8080/"
        }
      },
      "source": [
        "ANN_history.history.keys()"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "dict_keys(['loss', 'accuracy', 'val_loss', 'val_accuracy'])"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 113
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "QJkx2YfAVWEf",
        "outputId": "348523da-c769-436b-93e3-3325c99d9f55",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 404
        }
      },
      "source": [
        "\n",
        "fig, axs = plt.subplots(figsize=(8, 6))\n",
        "plt.plot(ANN_history.history['accuracy'], '-o', label='train')\n",
        "plt.plot(ANN_history.history['val_accuracy'], '-*', label='val')\n",
        "\n",
        "plt.ylabel(\"Accuracy score\")\n",
        "plt.xlabel(\"Epoch\")\n",
        "plt.title(\"Accuracy score vs. Epochs\")\n",
        "\n",
        "plt.legend()\n",
        "plt.show()"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "iVBORw0KGgoAAAANSUhEUgAAAfgAAAGDCAYAAADHzQJ9AAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADh0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uMy4yLjIsIGh0dHA6Ly9tYXRwbG90bGliLm9yZy+WH4yJAAAgAElEQVR4nOydeXxU1fn/38/s2ROSsO+CgIoLgqKItWLdtVSLYnGr9mv9dhHRrxbrUou2tT9b69LNLmqrVFAE3MXiXhUVGhQQUFBZwhYg+zrL+f1x7p17ZzKTBEggCef9eo2Zuduce4P5nGc5zyNKKQwGg8FgMHQvPAd6AAaDwWAwGNofI/AGg8FgMHRDjMAbDAaDwdANMQJvMBgMBkM3xAi8wWAwGAzdECPwBoPBYDB0Q4zAGwwGQzsgIoNFRImI70CPxWAAI/AGQzNE5E0RKReR4IEei2HvscS2VkRqXK+bD/S4DIb9hZlpGgwuRGQwMBGoBM4Hnt6P3+1TSkX21/d1FJ3sPo5SSq070IMwGA4ExoI3GBK5HFgCPAZc4d4hIgNEZL6IlInILhH5vWvf/4jIahGpFpFPRWSMtV2JyDDXcY+JyN3W+1NEZLOI/EREtgGPikiBiLxgfUe59b6/6/weIvKoiGyx9i+0tq8UkfNcx/lFZKeIHJN8gyJSZF23QkR2i8g7IuJp6R5FxCMit4nIBhHZISL/FJE8a5/tmr5aRDYCr1vbr7KeSbmILBKRQakeuIi8LCI/Str2sYhcIJrfWd9ZJSIrROSI1n+NLSMid4rIPBGZa/3O/isiR7n2j7I8ORUiskpEznftyxCR31rPolJE/iMiGa7LTxORjdbzv9V13nEistS6j+0ict++3ofB0BJG4A2GRC4HZluvM0SkF4CIeIEXgA3AYKAfMMfaNwW40zo3F23572rj9/UGegCDgGvQ/08+an0eCNQDv3cd/ziQCRwO9AR+Z23/J3Cp67izga1KqZIU33kjsBkoBnoBPwVUS/cIXGm9vg4MBbKTxgXwNWAU+rl907ruBdb3vAM8meYZPAlcYn8QkcOs+38ROB04GTgUyAMuou3PtjW+ifbQ9AD+BSy0JkZ+4HngVfQz/jEwW0RGWOf9BjgWONE692Yg5rruScAIYBJwh4iMsrY/ADyglMoFDgGeaqf7MBhSo5QyL/MyL6VA/2EOA0XW5zXADOv9CUAZ4Etx3iJgepprKmCY6/NjwN3W+1OAJiDUwpiOBsqt933QQlKQ4ri+QDWQa32eB9yc5pqzgGfd42rDPb4G/MD1eYT1rHzoyYAChrr2vwxc7frsAeqAQSmunQPU2vuAXwCPWO9PBT4DxgOePfx9KqAKqHC9zrD23QksSRrfVnR4ZiKwzf196EnIndZx9WjXf/L32c+hv2vbh8BU6/3bwM/tf1/mZV4d/TIWvMHgcAXwqlJqp/X5Xzhu+gHABpU6tjwAWL+X31mmlGqwP4hIpog8bLl/q9CikG9Z1wOA3Uqp8uSLKKW2AO8CF4pIPnAW2guRinuBdcCrIvKFiMxswz32RVv2NhvQ4t7LtW2T6/0g4AHLxV0B7AYE7RVIHns12lqfam26xB67Uup1tKfgD8AOEfmLiOSmua9UjFFK5btei1KNVykVQ3s1+lqvTdY29/32A4qAEC3/vre53tehvR0AV6M9EWtE5CMROXcP7sNg2GOMwBsM6Lgq2v37NRHZZsXEZwBHWbHZTcBASb0EahPa5ZqKOrRL3aZ30v7kdo43oq3j45V25Z5sD9H6nh6WgKfiH2g3/RTgfaVUaaqDlFLVSqkblVJD0eGEG0RkEi3f4xa0aNsMBCLA9jT3sgn4fpK4Ziil3ksz9ieBS0TkBLSAvuEa74NKqWOBw9ACeVOaa+wpA+w3Vg5Cf/R9bgEG2HkJFgOBUmAn0ED633dalFKfK6UuQbv9fw3ME5GsvR++wdAyRuANBs1kIIoWkaOt1yh07PhytKt1K3CPiGSJSEhEJljn/g34PxE51koKG+ZKKFsOfEdEvCJyJjpO3RI5aBdwhYj0AH5m71BKbUW7vv8oOhnPLyInu85dCIwBpqNj8ikRkXOtMQp6tUAU7fpv6R6fBGaIyBARyQZ+CcxNY+0D/Bm4RUQOt74zz8pVSMdL6AnELOu6Meu8cSJyvBUXr0WLayz9ZfaIY61EPh9wPdCITrD8AD0xu9l6xqcA5wFzrHE9AtwnIn2t3+sJ0oYllSJyqYgUW9eosDa3170YDM0wAm8waK4AHlVKbVRKbbNfaPfwNLQFfR4wDNiIdudeDKCUehodN/4XOg6+EJ18BVpsz0P/QZ9m7WuJ+4EMtKW4BHglaf9l6Nj3GmAHWpiwxlEPPAMMAea38B3DgcVADfA+8Eel1BtKqWi6e0SL2uPokMGXaKH9cbovUEotQFupc6xQw0p02CDd8Y3WmE9DP0ebXOCvQDnaTb4LHWJARH4qIi+3cJ8AH0viOvj7Xfuete6vHP1cL1BKhZVSTdZzOAv9e/gjcLlSao113v8BK4CP0KGHX9O2v6VnAqtEpAadcDfV+p0ZDB2CKJXsITQYDF0VEbkDOFQpdWmrBx/EiMid6CRD85wM3RZT6MZg6CZYLv2r0daowWA4yDEueoOhGyAi/4NObHtZKfX2gR6PwWA48BgXvcFgMBgM3RBjwRsMBoPB0A0xAm8wGAwGQzek2yTZFRUVqcGDBx/oYRgMBoPBsN9YtmzZTqVUcap93UbgBw8ezNKlSw/0MAwGg8Fg2G+IyIZ0+4yL3mAwGAyGbogReIPBYDAYuiFG4A0Gg8Fg6IZ0mxh8KsLhMJs3b6ahoaH1g7s4oVCI/v374/f7D/RQDAaDwdAJ6NYCv3nzZnJychg8eDC6cVb3RCnFrl272Lx5M0OGDDnQwzEYDAZDJ6Bbu+gbGhooLCzs1uIOICIUFhYeFJ4Kg8FgMLSNbi3wQLcXd5uD5T4NBoPB0DY6VOBF5EwRWSsi60RkZor9A0XkDREpEZFPRORsa/tgEakXkeXW688dOc6OpKKigj/+8Y97fN7ZZ59NRUVFB4zIYDAYDAcDHSbwIuIF/gCcBRwGXCIihyUddhvwlFLqGGAq4FbC9Uqpo63XtR01TjcLS0qZcM/rDJn5IhPueZ2FJaX7fM10Ah+JRFo876WXXiI/P3+fv99gMBgMBycdmWR3HLBOKfUFgIjMAb4JfOo6RgG51vs8YEsHjqdFFpaUcsv8FdSHowCUVtRzy/wVAEw+pt9eX3fmzJmsX7+eo48+Gr/fTygUoqCggDVr1vDZZ58xefJkNm3aRENDA9OnT+eaa64BnMp8NTU1nHXWWZx00km899579OvXj2effZaMjIx9v2mDwWDoZiwsKeXeRWvZUlFP3/wMbjpjxD79De/KdKTA90P3p7bZDByfdMydwKsi8mMgCzjNtW+IiJQAVcBtSql39mUwP39+FZ9uqUq7v2RjBU3RWMK2+nCUm+d9wpMfbkx5zmF9c/nZeYe3+L333HMPK1euZPny5bz55pucc845rFy5Mp7t/sgjj9CjRw/q6+sZN24cF154IYWFhQnX+Pzzz3nyySf561//ykUXXcQzzzzDpZde2pbbNhgMhoOGvTXUkicFXx9ZzBtryto8Seisk4oDvUzuEuAxpdRvReQE4HEROQLYCgxUSu0SkWOBhSJyuFIqQaFF5BrgGoCBAwfu00CSxb217XvLcccdl7CU7cEHH2TBggUAbNq0ic8//7yZwA8ZMoSjjz4agGOPPZavvvqqXcdkMBgMXZFkYa1risTF3aY+HOXGpz5mxtzlKcU31aTgiSWOUeeeJADNJgIvfLyVivpwwvHXz13Oz59fxTlH9olPFPIy/IhARV14v00COlLgS4EBrs/9rW1urgbOBFBKvS8iIaBIKbUDaLS2LxOR9cChQEI3GaXUX4C/AIwdO1a1NJjWLO0J97xOaUV9s+398jOY+/0TWjx3T8jKyoq/f/PNN1m8eDHvv/8+mZmZnHLKKSmXugWDwfh7r9dLfX3zcRoMBsO+srCklDufWxUXrIJMPz877/B2EaL2tnJvW7iC2Us2Yv/hT/X32yaqVPyYGXOXs3TDbu6ePBrQgp08KUimPhxlxtzl+LxCOOpcyz0RSKa8LpywP3kS0B4h4NboyCz6j4DhIjJERALoJLrnko7ZCEwCEJFRQAgoE5FiK0kPERkKDAe+6MCxctMZI8jwexO2Zfi93HTGiH26bk5ODtXV1Sn3VVZWUlBQQGZmJmvWrGHJkiX79F0Gg8GwtywsKeWmpz9OEKLyujA3zft4nxOObSu5tKIehSNwe3JddxL0qNtf5gmXuO8JCnhiyUYGW8nUpRX1FFPO3MAsikm/cklBXNyBNp3TEvXhKPcuWrtX57aVDrPglVIREfkRsAjwAo8opVaJyCxgqVLqOeBG4K8iMgP9/K5USikRORmYJSJhIAZcq5Ta3VFjBWcW1d5xlMLCQiZMmMARRxxBRkYGvXr1iu8788wz+fOf/8yoUaMYMWIE48eP36fvMhgMBy/7aiHfu2gt4VhzyQxHFXc+tyrttVv73oUlpdz41MdxK9qmPhyNu7Ir6sLNXNjuOHhehp/apkhcYOvD7RM6ta3+63wLGCdruc43n9sjV7Xp3LaeU0w5vw88xI+arqOMxJVRW1rwOrQHotTezIE6H2PHjlXJ/eBXr17NqFGjDtCI9j8H2/0aDAZNchzZLSqB/D5pxd4W55bc2zbua+4kHwUINLOi/R4hO+SLi7ZbmA8kqYR2TfAKQhJudmyD8jOy8R8pr7M2eDlBab7M2T4n+Xvu8j3CNO9iylQ+5zb9MkHk++Vn8O7MU/fpvkRkmVJqbKp93b6SncFgMHQVUtXiaEt9juQ4stu6tOPOg5POd7vN24L7mrZcp5LtcExRXhdGoePO+yru++oKt7nB93R8/DYTG+/nzeiR8c/1KsCCyAQmNj6Q9jrPR09AKbCdHY3Kx4LIBCY3zmJuYBa3+p5gnKxlSfCHfBX6Dpf5FuMR6OWp4KPQD1gTvAJonxBwaxgLvhtxsN2vwdCZ2FMXeXJCW6bfQzimEgTR7xEQmm3LDvkorwvjFUlwfae3SH2MbPxn/Hy/V6hro5t7b6zcZFpyU7e0D7As4NeYHZ3UZvd5W8c/sfF+3gzOIEuaiCld7jvV9xRTzgfBH+GR5nqpFDwRPY1LvK/jk7Y90wbl55XJH7dLgp2x4A0Gg6EDSZVENmPucm5buCLl8bctXMH1c5cnJLTVhWPNrN1kwbe3ldfp85Lj2hMb7+f5yPHYmyPKg1La6nSf31Zxt6+5MHIijUqnbNkWa0tWbjJu67+YcuYH7mB+4A6KqUjY52ZN8AqXBay4zLeYr0LfYU3wciC1ZZ9q28TG+3knekT8s9tKv843n0ya9HYCPBGdRLFUphw/KLbEClLe32W+xW0S94jy8FzsJN44c/F+WSdvLPhuxMF2vwbD/sAdp7Yt5n5JFnq6ZbY2+a4EsrwMf4KwtzcP+37LGb5lKfelixNDeku6mHJeCN5KMRV4RFusz0QnMsBTltbqtklnPduyk6pHlnuMt/pmc7b3AwISJaq0RTovOpGbIv8bt+z/FZ1E9oUPAhB74QYmRxY1s8KXBa+hUGr0MUqIISkFuVH5GGF5Oloaf6r7UYAnTc8vpfS9KuDLwVMZeuXDrV6zrbRkwR/oQjcGg8FwwGjJrZ7sQofE9dR2BvjPzju81Wxo9zU6UtwBBnp2AFqsfETxiqJJ+Xgxejy/jEwDbIt6DS8EfhpP/LrONz8hK9wW/I2xYnpSwW6yKaSGjaqY4zxr6C8702aQ2+de1fh/PBj8PXnU4RcnRyCdsL8cPS4+xjIKqCEDP/o8r3XOFN87TPE5hU0v9S2GZ11tTkRb1Jf5FseF1Y1HFEopFkZOjE8empSXgES5MZzY9uQ8zx/4Td7THFnxbwQr7q6jJlaCoSBWJoLgTFzsY5R1fF1GH7IatiKHnMZQf02a31z7YwTeYDAcdKQSb3tt9tINu5tVJ0tHeV2Y6+cu78ih7jFr1EBGsYmgROKC4yNCDRm8E5yeYJH2Ep345SZZHI/3rAGgEC1MgzxlzY51W90PB37HANlBIdXcFXiUQlLXAQErSc/6ngBhashI8AgUSSUN+KiJZVIkVdoKTrb+fRkwbBIgsPZFUDEi3hCvqeP4Xd1Z3J4xjwlqWcJ3zo9OpJEAPmvy4ENb8zf5nuKDpsMSVx4seAPb4+8RwBeESBNa3lXiWNyIFyEGxSPIKhgEn22FoSfDhOlpn0d7Y2LwnYzs7OwDPQSDoduysKSUo3/+arP4t019OMoTSza2m5Xd1jhxe1FMOWd6PqRS6WZUtvCsV30plkomNt5PVKX2I9vCmc597o7m2tdIzjp/P/hjxnjWUSxVeERxiGcbIs61kgPC4vqezaqIc73vxWPzAH+JnEOGRCj2VMePc18PgEg9ZPeCrGJQWqh90UbOGDOcV+75ARMO6dHsO6f43uFS32vxbU9Gv05UCQM9O/jopGW8O/NUJg/zwqNnwc7kYjQCY6+CyxZC3oCkXV7I7g1HfQe+/5Y+rnAYfOcpyCyEXevZnxiBT6Z6m/6lVm8/0CMxGLoN7dGKOd01Wrq2LeiDZ77IuJlP0GfBBfjrtQXakUJr43aF29+TLqksmT2ZHNjbb/XNJkPCbFWJ/Sx2qnyuDc+gjAIWRifomLErEW8DvZ2DBegxlAi6sqf72Ealt3mtbPIgYaJ4+MBaEpYu0SyqhHrlB+US+VCec0Buf3p6KsmnjmNkXfzZ3Oh7Wu/P6gn5KfqN+LK0mNZsh2pXM9LDztfbAKqs7UO+BmJJnjfApv7nMs97DkrBNN/reEUhAEv/DnfmwX2jYOMSkCRH94k/hnPvg0O+DsO+kbhPRWHkOfCtP0Hv0fq4qbP1vh6HwO4OLcjaDCPwybz1//Qv9a1ft8vlZs6cyR/+8If45zvvvJO7776bSZMmMWbMGEaPHs2zzz7bLt9lMHRG9rVMqdvqTs5SHzzzRWYkbb9v/tvsfGgSL7+/PKH0arKw7ovQtkZiBrizBjpVVvgXwWkpr31j0rptO9kt1ZiXBH/M8Z41TPa9B8BIz+b4viqVwQArLg+QJY3aIW1ZwT6JMSjbsYpl8ESIRfASdYQdH1GE8nh3b82uorGc6FsDqLj429gTA6XAgyKKh8ejp3Fl4D5t6Ta4stWrNhMiHLfO7WczwWt1F6/dDhUp6r6rCJz2M6jfDcd+19l+9DRHWIecDP4s6DHU2R9tYkDvXlw0th8SyoUjpugxJVw7pl+bP0zcvuVj533tDsgfBId/Cw6brN/XpDEOCw+Bss/2qwF58GTRvzwTtqVesgLAxncTfVA2IjBwQupzeo+Gs+5pcVwlJSVcf/31vPXWWwAcdthhLFq0iLy8PHJzc9m5cyfjx4/n888/R0TIzs6mpmbvkjBMFr2hM9JShnlyNjqkjo/vCXf5HmGa7zUWes/ghtrL25wJ7V7XbSeJ/azpCv4RvIdiquKZ2W3JQO9JOe8HfxS3dN3Yse0G5WeLKmSwZzuzI07Wd1vH675WW7Dvz0OML0KXsjvvCHqoci1itWUQs6qznXo7bHgP1r/G5j6ns3vbBo5Uba+ZbieWoaAJLwL4iPKF6sNpTb9FgC9vPRYW3Qorn9EHJtXEUwrqJUgGjXoe4suAzB76ZvuP0wesfwMaK2HMlVDyT+h7LJR+pC9w/u9hzGX6/b+m6slBjyHalb/tEyhbCx6ftupLl2pLfNmjzhjyB6aeULjxBeG2HS0f4+bte+H1u/U9HHuVtu7bAZNF3xb6joPyL6F+l/4HLx4dMykY0vq5LXDMMcewY8cOtmzZQllZGQUFBfTu3ZsZM2bw9ttv4/F4KC0tZfv27fTu3bv1CxoMnYi2FHdpKcPctsSfXrqRr3bVt7mqmhtbXI+WdQklRC+IvsIFoVdoUD4rY/pDAhKhQfnYpXLpK7sR0dnmL1kZ5u4ksSKqeCV0S/x6dkJZRAke4DrffB6MfCueaW5b1rPU1Zw+7ije+PgETlPvxUU4ojxsUT0YIDsBCEmYobIt4dp28ZXbfE/wTd/7Ld53FFilBqM8IY5Sa5ofIB5QMZrwESDCd5p+CkCGbtTJ7Nqx/Dj6uHYru3n9Lmzzvn9WjP5TZ8GTF+t9A0+ALSUQaSB1oVp7wiFslmJqY35GeLSn5q+RcwHom58BOb0hmKsP9gat6yVeI9Map77ZRjj0zERRnGWFIf77mP5piztoy9qm/CttvdsW/X/uh80f6fFv+S8EsvXxQ06GL9+C0VPg0+S+aM7zxJcBo86F03/R/Jh03N0TItb9KKXDAEv/vueThD3k4BH4VixtAJ6fof+x+EIQbYJR57fLLGvKlCnMmzePbdu2cfHFFzN79mzKyspYtmwZfr+fwYMHp2wTazB0Zux2nUWUMyfwED+quI7r59Zzw1PLiSltnX99ZDGepGprySjg3fV730vKdrXPj55EvtRwmrcE0Algr0TH8XDkXP4R/DV+tPgHiJItzkTCT4QoHn4feJCxsjal1e3GZ+23RRmcTPPLfIu5jMWwMggDxsBGx8L2Sox+ssv2jNPkzyUQrkoY6y8j0ygj34l/I/r4HkOaxW+9vhBHRjZARiHUuXaE8iC7D+xcA74QvnAjCPzS/wiXNd2CLcrXhh8HSdcm1XoG6xbrl01mkf7b6As1E2VAu7mPuABO/wVL10VYPP8Rfs9vANiiChPLs9bu0G71sd+FOdO02zq7CCpLSZg4nHMfbF/V3PU9+c8w/3uph//aLB1mvXW7Fvhhk/R2t9CioGKDflv+BUx7Rgv8iqfTPBIrvyBSrycnOb1SH5eK6Z/Asz90nuXeTBL2AhODd2P/g/veYv0zXSxlD7n44ouZM2cO8+bNY8qUKVRWVtKzZ0/8fj9vvPEGGzZsaJfvMRjag7YkxC0sKY334navnwanRrfdL7slcbdpLaEs1f7kSmff9r0TF/eI8hC0ll1N871GTyqoJBPQ66DzpD4uvB6BC73vcLxnTavi3iZGT4HpK3RMFnQSmCXUXpdw2eKuID5W2+XfR/SERzILdSZ21PZMuPzxkQZ9dt1OZ1sgW7+KhsHYq+F7i3nFewoAh8omrvPNJ0u0MD/ovwoOv6D5+L0BJxnNl6HvR/z6c0OF/tuo0lRtU9G4+E0+ph+Tv/G1+K6srFx+dcFox8MzdbY2oHqPhus/gdu3w7DTm8cctq9KTFaz6X0EqRGd3T59hQ7LRuq1ex+00B5xYeKx2b30sXn99aY+R7t2e51nIV7tyreT+vaEnN6Q209/ny+oPRJ7OknYCw4eC74tuP8BtVN8BODwww+nurqafv360adPH6ZNm8Z5553H6NGjGTt2LCNHjmy37zIY2kpyrLsg0885R/bhmWWl8cYltgv9+rnLOTKvnkey/0TRlf/i3kVrWZ0UK05eE23jjmf/PPCPlNXPZvjmMU7W8hPfk/EKaclJcMmtOSc23s99/j8x0bsS0FZwDMiSJt6OjeZkz4q4hQ2Qb5m5MaUtY7eYp4pj25FhNzHlOKbdS7wS+NIqwrJ9pQ7zrbLjzKkRYNOAcxmwpYyekXIezvgD0aMvg5JbdPLY2ffCExdC5UbweJ1YeSqaanSW9ykz9ee7e3J2TFusHkn0OkwYcyRES6xBeLVoB7Ohqc4KU3odIQpmQ0O5XvJ17n3wtZvhpZthzQta1MULh5yql6q5xO+0E8fD6/r9n0evhdbKs9buoNlTT+fOzunjvM/qqc+13ejRiBbPF2/Q+zcusc7pDcE81z1H9SQmp5dj2dda6/w9fohZ/75tr27+wL3XhrpderI29ruw9NF2MyBb4uBJsjsIONju19AyrVVpu+npj1P2/06HXRp0nnyDmxuujJcSPcf7AX6JJlQicwu4fd7nqi/DZUtCGdE9SShzY08i5vhnMd67hqjSFnIVmeRLLctiw7m26Xru9/+RCd5VAMTQLssVaghH8CUxBA+KKjLJkgZ8xOIyHJeYUL4WgJqtgFOStJIs8qhNW5qUsVfDpg91otaVL8K7v4OV80kr9Fe+BIMnwAs36GSvgSfAhnf1vhvWQNkaeHwyXPQ4fPqsKznNwpehLVWAC/8Oo7+t31dvg0W3EVs1H4/SWfE7pJBe7ILLn4UP/6otWFt01r+h3dUAE2bAew82j9GDFtujpumQpjegxe/Y7yaKX4I7POncluLO1dtg0U91HDwWTnRnuy1epeCX/SBcC4XDdQx97HdhzqVQ8VXqa/uC2gr/4k0YeKJ+xpkFMEP/G+E3h2qBVzGYOgde+YnePnW2I8rJnoQDjEmyMxg6OXvaiayl84CUVdpuevpjfv78Kirqwq3Gxd0ki/BFvMpFoVdpUH7mRU+OVwMLWNXSQDE3MKtZ0tsI0a7+5ISyX/v/yqne9NXgIkricW93XH1uYBa92QVAFSGiykeRp5qYgoGygzIKyJVaQNcfF+sag7MVtbUZfBg9lOO8a8n0KnxR7XIWgILB2k27pUS7jwEyCqDkcVargRzm2UimauTD2AjGe9cSUZ7m67+X/t15v+xRy2pUjqWc0ydx3fY/v+lYi+CIO+jj7IzuvkdrEU6eKERcyYkZroYoOb3BEnfQHode1jNj9hS43alKx7n3wXu/h1dv1Z9HnAXjr4VFt8Ga53VIwC22L97gxNBTWaTTP9HnrprvWMptiTvbVraKWpZzGne2CGQXQ3ktnDQDjtElbhl+GnzylBb8NS/obcmThL+cop+h1wcjznGumddf34c3CIeeDiPPSnw+XQwj8AbDAcZeJ+52i98yXy/pbK3daPJ5LZVNbakLWTLuZV8TG+9PsITdCWF3+R9lverDMNnKG7GjKJbKhKQ3H9F4vW87m7xe+XnFZek3KaegCjR3e9viHsOJVX/X9wrjZC2lVkGXAqmPS952VUAfTzkD2M5I0cIYFU+8pnnOyK/DinmcesyJsMkDlZuBbMgfDH2O1H/gL2reBrWyZAEe61uCEuG3kYt52jtLtxD1WnHVVNhiLx5d3ezhryWKO2hxFw8gjsvbtpyrt2mBFy/k9NWu6CMv0bH39a9Zmd2upG/Oh0IAACAASURBVLdVC5ykMtAFWXZ/AeUbrGtaQYbLFjQfa45rJU92T0tsc5zEOrfYthbStM9F7Xnc2Z2A15I7O2w98y/fdgQ+qyc0VkHpf62DpPl3ZxTo8EdjjQ4/2OT2g9JlUHSoDod0cYzAGwwHmHsXrY2LtE19OMq9i9YmuNSTLfVU57UX7vj37ZGriFj5uGFX8loZujraw/77GMZWTvasSGgo8m2rIYh7LqEUhAgTCAQ4Y+yRPPnBJgpF1yq364y7hV4BYomXBw94hMvEiavbS87s4wH6eMoBeDN4Ax4UDcqfGAao3gZN1ZBVCBk9ANFu2aJhLVppsfzBjKxYGf/c1/rurb1Ppd/kn8O87+nM9WSSrccbPoVXbtFCbE9LCobAgPHwyZPWg4o6k4aqLVC5SYuP1+cI6/MznO9wZ7SXPK5ftiv80mf0sbsfTRxXqspw2S7xfe8h/TzaKrap2NtzW5s8JLv/P5mjX74gxOyJkT2Jsv5RLXvUuVZGD9ixWj/ngEvgM6xkvPxBbRtnJ6fbC7xSCmlrJYguTHfJpTgYsdeJJxdLKa2oZ/DMF8nP8FPbFIn3Bbct/H0R93StQZPd8bY73Q7VvxQ9nkqy6OnqmZ1nucEfikzmW97/MFi2JzQF2a2yKZQadgf7Utik/+iek/sF50wezdhBPXh8/rmMQ//hFYFtsXx6eSqcTl2DJmhLlZheyrTsn7DaVf0xkAVNtfGPdgKcnUQXwiXuwTzY9bl+n1lk1Qdfp0U/p+U6FAX9DgWXwI/KrIII9PvWXdDrcD1BELRHwBfUSVWpLNec3jq2714DHmlMzIYHLTwNEceCTxZkWzxHnafd5bu/1HefyhVeuwMGnwRfvaPHun1lorBBc9FMldy2p27qDkpcjrv/U4UO1v1bL0mzSfU8MntAtc6r0F4Gi7LV+qe9r4vTrZfJhUIhdu3a1e3FTynFrl27CIVCB3oohr2gb75uDOK2mt3Lwirqw3Fxt9kTcU+1xCxdmdaJjfezMHJi3G3eqHwsiEzgdxGdtBWQCHdEruLasGM95qHF9Qb/MwzxbG/WFKTQoysz2uIO6PXHd+Yx+cVjuN/3YMIYensqrAx1689Tleu8l29OjDeDLkPqotl03uf6/6J4pF4XDZBVpP/QV27SLvLsVgpNFQxO+Hjt0QH9JmiVb506G8Z9T2eyx6J6MvG911IvuXUvyc3rDzXbnGV1NsO/oS3q6q1avHeuTSxxai8zO+TrMOQU/bDTxaynzobxVtc4u2RrIPG5Mf0TXXLVxl4iN72FCqAHinShgwdGJ4o76H8vK59JfB4ZrgY0wRw9ubkzDzZ9oLdtWaY/392z4++lA+nWFnz//v3ZvHkzZWVlrR/cxQmFQvTv3/9AD8OQgtYS4dYEryAUam41K0XKftvprO90uMV8ivetFpe2Jffg9luJc3bsuac0r5me79EC/0WsF4NlOx7RiXFelBZ78epM62GnwdqXmiVcef77OLxxly4dGos4+4aeCs/+LwwY51hWu9ZDfWXiABoqtUu13xjtNthS4hQwQbQI2DHt+l3OMjPbgrfXdLdiwTsCb8WwK636ACFXffZia8lrQ4UWbLvhSDJuyzaUr63+9x7S187ooceZ3VNPQlY/r2PKoIu3pLpeW1zhdvJd5Sb9rL2BxP05vSFUoMfgC+y3tdp7Tap7nv5JYglcb1BPoNx16MFZFw/ak9HMIxDSnpEOLkTT0XRrgff7/QwZsm+lZg2GZlRvg3nfhW8/lvKPn1vQ85Lc600VW+iz4BZ+1HQdFZY4T2y8n1t9sznXuyQhG9tuupG8ttxeM55K/N2kcreDFl+FB79EaVIeqlQ2lzY5JVn7S1nczV0SG0axVBIIZkAU+ngcgS/I9POz8w6nz8uN0KQbigi6c5hd0CWK4FVRbS262nkSaXDEw15rZieL2cJSaP1RLnk88caSXdnRRi3oNdusa7uT3VRiEsCudc77rKLmGectERd463qVm/VTCrhcvD1dy1Rbi+Mmu8Tt+H29leUeyLL6Z7jGn25NeFtc4baoVWzS104Vuqzdsd/Xau816e7ZXQI32gRDT2n+TBIs+OwUHoGmzj25aSPd2kVvMHQILXQcTO6cluxeT+Uat61mr+7xlZBo5u63bVdu+47vjYRuZGuCV6R0w5/S9ADPRcY3u9ZCK7sdwE+MQqlimqs3dhO+uOauD4zkzLsWc/FoXRykr7eSr351Nl/dcw4ld5zO5CN7apc0kCP1RBHmRyeyURXTiJ+anGHaShbR4jHmcm05Fg7TSWbV2534eXIVSfuPcGaxtsA9tj3iSfzsdiVP/0R3BrOtU29IVzXzBa1TXTZNZqF+2bQk8Hf3hH+en7htxypAgcf1ZzSryAkJ7FiV/nrgjDUdb/6KlOvm99Ztbk9m6nY2j7/buKvLpaoe1xVoS0XSTNfEzp6gdVAl0wNJt7bgDYZ2JU0SUiN+xnv+RUVdmF5SwWP+B/kRbUtesy3zIqkkggc/MbaoHvTz7CamJCFjPVXlNnu5mrtc7IORb/FI6H76FWSyelc0nvBmX6u313Fx20acu8KZm4uiL+j7Hn663hBtgvpyxxp0tfzMyMjC11DNKjWE+zNn8GTRowysXg7ebBhwPFz4V31gzXb46j8QrtOTJI9P10+3rSz7Z40VWqvdAf5MCNux9xh4MyHakDqJLZhjufotS8yfBdGwtZTM+v2JV7vG3a7almLwtuv304WJleSyXdXUkv99rFus47jpCrvEl5FZyQrpyr/aiMf6Re6lZRlyhXPSCXx3oC3ejGQLvq3ndTGMBW8wtBXL4oqJnhfbCWgnNTxAeV0YBfzQO7/F5LVGpc9tUH4WRCYwuXEWcwOzeCj8LfyWez5m/W+5URXzRHQSxVbGehkFBGkCtBs8SJjzvO/zUegHCT3GPwr9kCP4nB7lH3OcX1clE4HXvCcyaQB8485/I/6MxE6J3hAN/nzekzFElf5+hdVRcfoKaKhyjq3e5ryvtzwGHj+ZDXr7nZeezrszT2XgwMHabV632xHSu3vCZ69oq1/F9CTpw4cTe4PbuN3nHp+2/rHXJscsaytFEluyJWbXT//eYu2uBX1fHo9jwYfywd9Ckqrd/UzFHG8AJFqCtkWeyrOQDtslPuq85vvi/cmtP9OjztfH7q1l6Qs41mpygt3BhvvfVjee7BgL3mBIQbrEOLW6ismxCIiTgFZGfqsWOsAJ3k8JxDuaReKNUMbJWmb458XP9VpCnyUN3JEUY7ebkFSQzYvR4xkgZVSqLM73vq8LrljY0VV/1Gkz9o2CbXD+P+Bvp+nM4pzeukWyeCHaQCjawIlFdbAzhm6OEtMWaU4vneQVyNHrx6u3Qq/D9EUbLIEvGOTEt/Os4jw5vbUFHW1yhHT6J/DcdfD5Iv3Zl6H3uUUz/iB8WlQbq3QFt/MfhM8W6UlDryOaW/w2yZaYe/+E6fDFG9qVDo4lF23U4YKWLGN3UtfDX3Maq9jYFrl7DXtr1rY91jnTdHnbup26FK149HWKR8KFf3Pi4ftqWWYU6N/hwS7wmUlZ9N0UY8EbDEnYcfSmii3MCcyiqWIrNz39MTfN+5iMpt2UoePR61XfuHVtW+gRy/q1LfSJjQ8AOvbu7mimcCYBHlHxLmhKQV9LxHtQrUXWQoAKj3azFko190am8t3wT6gjqNeLt8au9fDwSbpSl/1lkFhr3F2opddoq/GI0hZ80XC9/eWbneVatgXvzlLOtVZzuIum2BZTTm8n1u3xWdXflHbRp8I+76v/aHe3nUhXunTvljHFM8k363uw/9CH61PmVCTgjk9nFett7gx6sCYBV8H/pFke19q1Y1Et9N9/W/8sHNa+8fAMy03fja3WNhHMdTwt3fhZGIE3GFwsLCnlxqc+pj4cTUiIC8cU4aji2vD1hCw3+U6VH18PnpwoFyBMFA8fBH8Yb2kqAvlidTRDElz2Ct2pbKk6ND4Wryjy0Qls/fIz+PKWoxnNejbFtDV8iOj14VO9b6buapYWS9g3WR22snrqzlmgLU9fhu7pffR3gJiOuTdW6fKdoCcKthg2JAm8N+hYx26Bd1tMdbt0XPyQSVoEw3XpLUr7vNFTtPvbTmDzhfYu2cwW+MYquG8U/MIVd1/697ZPGmxhDyYJ/L4mqXV0kpv9PA92C17Eyrb3ON3juiHGRW/otqRt4JJimZu7dera4OUJTVIcd7uPT9UgcqWesPIy0JNonRVJJWG8BIjyqRrEcZ41gGJDrJhBHv1HJKb0qrDfRqbQX3bGXfZ269FDc2NQTbxuew+ppsFXoEME/74DUNR6coFd/Nb/J6Y23c4fI+fyI//zNOGLX4/c/voClZvSP6AeQ/VSrq0fO41Ook36j15uX8fSrt6m/wh+Msc6UTnLtWwryI7n5/Z1MvfcWenupKaps+HRc3TcfdpTuutaOjepLch5/aGufN+WMSUnwSV3SWtrMxRwhL2ruXft53mwCzxob4mKpa8t0A0wAm/oXlRvY+dj07h41zWsb3Bcb3Zf86UbdnO371HYuIQvnrmDCzZcmNB1DeD56Alc6H0nrlN2tnoEDxd6dX31/8YOYZznc4I00YhekvXz8BWcGdIti4/wbIhfb5BocVfKiY17UBRJVbOya3nVVhzb2j4yp4FXI5fjfdYRppF8CcBQ2cp1vvmc4FmFUuAXV+b4oWdo9/Dw0524bjK7v9AvgB6HwO71MHA8bHxfC4Hdb3ved/XPXkfAdmtpmC9DNzTZUgJVpc4a8bpdTiw722UJuy140Mevs7L2m2rSC449McgogLK1e18THZoXQfFlQG4fXSVuT5uh2BZ8sou+sxMX+O7rlm6VtpTk7SYYF72hW/HBYz+hx85lXBl5qtm+1cEruHv5Sfp/ZhVj6FdzWK6msCZ4BUB8nfm3fe8kuLxDNPEt37tM8b0TXx9+vPczPKLoL457b7hnMwBfxXoSUc4Fotb7GoLx6+ZJLb/lMv2hxyHg9SeM1T77D02365rqI89tdj92IZxhnq26LCw4TUfsvtXuuG7/4xMtTvFA8QjtnrcT49ztUW0LvGyt/ml3BgMthrvWaXEXr3NsY5Xjvg/mahGFRAsetMDXbNMx/qaa9IJjC1JGj313X9uZ8O6SrrGozkzf07XP9nNMdtF3djKMiz6+2sFepdCZS/LuI8aCN3QPrFn58QBpKsAlV4xzryN37z/H+wF+q73pBtWTFbEhnOv9IGWc+9+Bmzi+8Y+Ukc8xopuYvBQ9nh/4n48f47U91uJYDX2DjfyjeCGyDS3KsWjqlqN9x8B5D8LfTk36ZqtcqrutaHLXMhu3ED4/A/77mC4EY7u5Y2EdVwfYaTVieelmV39yS9Q3vu98r4pBmZWQp6Lwl6853+G2iDKL9CTAvXYcoIfl0q/YqAvdpBMcu2CNtJMtkqq8abps/JawhT1dcmBnxbjo962NbRfDWPCG7sH0T3jTMz7+0V0Bzq7yBpKQCOcuIgNOopyPaDzBvI4gP45M58XocYCTeG5nywPx5jDf972AUiSIu6Z5hvt50cX03fa6/lBVahU5iUGRVcvcXja2pURnvkebUl/TFnePv21/qJLXiEfDzhjAWep2wV+bV1kbNAGu/rd+H1+jbRHIIu53cFtE9vXfS2woE3fpP32FlWSXxoIv1SEP1ryQ/p72hPZKYrOFvctZ8JbAL/tHYuOag434/wd7uNqhi2EseEOXx06mm9mkwEdCBThQvBC8lWIqeCHwU1apQTThJUiU/1p11t0USRXbyadM5TFctpCNdnnbLVGV9R+7ZrxH0leBi1MwRK83j38eCo3VUL+7WeMVXrwBBk/QFuafTyLV5ACAfuOg9COrMclu+OYfdCes1v5QJa8R/+xV+JdLyO2kvMJDmieQhfKh/1g93uSObq52rUQbYeV8WPG0s23pI/plxzltgY+7/1tpXbrmhZarwu13rAlea9XnOht2LkT5l906uaxVumHVulQYgTd0StIVmkm1ze6N3tOvl2zFgKejp/Ad72sJwttLKujlqtVeSwYFUk0xFXEr/trwDP4TvI51qj+FUsuuwmPpV5/BQ5UXMNG7isWxMbwQOp8bw39moNrW8vI0uztaU41OWKvdqd3esbAWdhXVrme35e3+w5M8MbA56hI4air885s6KW7tS7q16FEX7/mDzipMvT2jQFs5fY/RXgSAz16Gn7fQva7HIXDOfbD6Od38JVSg30cbEycxCeJtTWBevQ1ev8sR73h3rxf0ZGJPMtz3B5uttqKrn4OjLzmwY2kryc+9GyeXGTTGRW/odCQ3bCmtqI8XmknedsNTy+O90R+JngWATxQXe9/QyWlpiCkY7/m0WVlZL1F6s5tSVUSdCrC7sop3Z57KU5dr1/npP7ifB2+dQWzw1+LXcTcri78tHAbn/k6/ry3T1d/smHblJm11Axx6VnoXoXuZmXgB0ZXNGqudZWlfvKXdxe7Sm3tCZpHz3i7eAvp6U2fD129ztn1vcWIjF9tN7w3qicrQU+CQU7RFdOkzVk34cPMe5fEmK67Z0YDjE5Oc4t29GtP3OD8QJPcNX/tS1+kbPv0TOPzCPSula+jSGIE3dDruXbSW+nA0oUOaXWjGTTimiLk2FUhN/L0Am+kZF9x40Tbrp0cgKJFmXdl6UY5PYpSqIhoIOA1O6sv1TysLeUioluqcIVRILtuUFtcYgtgiXDwSBp7oDC67l6ujmdcRxz5Hpo8F24lQPQ6B77+ls70Lh+lj8/rruHu4Vq9n37NKNw5ZLoHvdbg1Pr/z3f2PdY0nJ7GRi11KNV3VtnTdudxJTjbBnObi3Rm7e9mTk3jBnS4kkjm99WQwuS3vgZ40GToM46I3dDq2VGhRtSvJ/cT3JEM9WwH4ftMNCV3a3NzteyT+XgT6o92ObpG3O6ttUsUMtIrPuLPpB4suhWoLfK7fSmKzLW7bUp46m9zXfwFv3ws9h8POcjzfWwzL/+UsUXM3aAlka0veF3KWsoEW6nRUWPHw/IFOUpiNx+vUf3eXid1TAlm6pWq0wfIKvKnv0Z4wZBRAMA8aK2HJn1JnoSePzaalOGftDm2125ZwUw3N6Ixx0mZ9w7uYSKb6/Rm6LUbgDZ2ONaErCOIUn/m2TxeXUUpnrN+e1IDF5l/RU/mO9zWieOOV6CJ48A04jtimJXF3lQgMdBWfCUkTg4pzqNlZSL+obsVaqooIS4BD86yz6ssTLVvQVdtQWqQbKnUCWv+xzv5gjtXm1Krnbv9hffH/nDKxuf2a30hygtkXb6ROMMvpqwU+ex/dw14vRIGdn+nP9iQmeRzLrAmUL9i8icueMnW2Di/YPdbP+OXeX2t/05VFsjNOmgwdhnHRGw4IC0tKmXDP6wyZ+SIT7nmdhSV6mdZtC1dwUoNu3KKSQuh2YRfbnZ5MkAheVEKZWR8x2LQEJQFejJ0Qv2ZUCbtUTrxAzBi1ml9dMJpjM7SnID8nh8F9iikM2ha81QPd7QrP7at/bvpIW9nJiDj12I+9wlmedeyVzjF5A5qf11Y3cN1O/XPbqubXaAt2PNnOgN/wrv5pC31Hu6N7HeG870qV1Tq6XrzB0E4YgTfsH6q3waNnQfX2lEl0t8xfwW0LVzB7ycb4evTk+Dno9ecvR8bFu7S54/QFUk0NIb6I9aLBauISxQOjp+C9YSWHDR2IEqFB+fGIolCqnQtXbGDys4dxWXQBAM8c+RF9CvMhbLnT63Y3T2SzBb6pWtd0T4Ut8G4rO8MVYshLYcG35ga2hXnHp/rzV2/vXaJXckUvr5UjEMq1ys12sDs6q1A3ugHt5TAYDO2KEXjD/uGt/wcbl8Bbv44n0bmpD0d5YsnGuKgXSRWfqf40Ki0+tsh7ibGTvHgcfoZvHuNkDS8EfkqxVLAyNpT3YkcQIKqFHBUXpSGhWjxjryL0v28gR16iXdzJxeBtlv4dVi2AXVZlt/ryFALvEudUFjw4BVE8vubbMgvBn5H6vJYSzNrLso73L49ZVfSsYjoNlU652Y5OdLNL9H74t/a9rsFgMDF4QweTorHDu/ydhqBTQjYV14Zn8Bf/b+kru9imsqlQWRzl/YpSVUixVDbr+GavcY8qoTKWxQLfGRRMvIZTa150RMntSr3gz7ps67JHnW15A5xCL3YjErvfeX1FcxHPKHDKyxakseArN+qfq5+HIy/S70OWBR9ucBqzJNNSrLQ9LevaHTo7v+RxErLaU62Rbs+YbfK/i+VP6JdZk20wtBvGgjd0LNM/gSMudD77MljkOTmhhGyxq/iMm55STklsOF9reoA/Rr8FwDXhG5npu5nXPBOaxehB91A/xbOcC2+fy6mnTGo5Rlq7Q1u+NvZEwC776s90RKg+hYtexHG9J5csjbvRV+vPq59z3Oi2iz5c61jKe0p7WdZ2PPn6FXDEt52a7x29/KsrLzczGLoIRuAN7U5CAt0fPmXzLmcJlIo0UBkLUUZ+fBmcXWjGFvyRbGBuYBZ9ZDfbrTXmO5UW0IWB21mupnC2eqvZ0m+lYG2sH1NCD7dtoOv+DStcXedsF3UsokUz0qCrqClluehTLM+za62veSlxezoBU8B9o5zjlv597+Ln7Z3oZXdag/2z/KurLzczGLoAxkVvaBfs0rKlFfV2vzFAJ9BV1H9Gf0uMq1QGF/JvLgq9Gj/XruUeUYIHeCDwe4bLFkDRs98g+lVmsLtSi8/qI3/CUepzWDUfVJSY1WNdRL/qyOTqM8fTJuxyqCufTtphlfH0+HR8uqlGJ4G5e5onu5hXPq1ftos5nYBdb/Uk/3SB7iDXmUqw7u/lX115uZnB0AXoUIEXkTOBBwAv8Del1D1J+wcC/wDyrWNmKqVesvbdAlyNXqF7nVJqUUeO1bD32FnxduJcsud8SXQkw7y6V3oudTwTncgYz2cMke3xwjMiusQswAgpjZ97yvZ/8q5vLvzsc7jnRo7qkwm7cuJ13IUY2yiiD3rJ2OB8P8cckyIzPRXximpivazGIbboFg6DN38FVVv0dreL3p4crH7WEvEUQp1KwGxLWanOZ7nu7zXSZk22wdChdJjAi4gX+APwDWAz8JGIPKeU+tR12G3AU0qpP4nIYcBLwGDr/VTgcKAvsFhEDlVKJaZeGzoFqbLi3VzlfQWPqy68XbjGRkTXdI8i+EXFBR/Q5V6nPKZF0Btw6roDHHslgtCnYqN2twMFZ92+Z4O3k8zqdsKnzyY2f7Frs2+z4sLunuT25MAu25pKqNMJmLFcDQbDfqAjLfjjgHVKqS8ARGQO8E3ALfAKsLOT8gDLVOKbwBylVCPwpYiss673fgeO17CX2KVlU6MI4yGgosQQvKIS3OpKQRUZ1KkgvaVC73PH1jOLHNHMKtYd2SbeqDubDTsNRp6j9/1mJNRs3fPB2yI8ZxqMvTpRdO2qdMse0z/XvpJYpGZvhdpYrgaDYT/QkQLfD9jk+rwZOD7pmDuBV0Xkx0AWcJrr3CVJ57bR72rYr1Rv4+nQXfxvw4+b1YhfE7yCkDglZ72W897jEnARyFX1NOFHBJr8eQQzsqHKctPbTV5AN0apLYOvLA9ATh9nn8/qcLZyHow8e8/vI5XozrIasdjf99nLiSVjjVAbDIZOzIHOor8EeEwp1R84G3hcRNo8JhG5RkSWisjSsrKyDhukIT1fPPMzxqg1CS1XbSY23s/HMacRSkQJO2M58eVtDcoXd8cXe3RjlmC40hF3gKJhzvusYi3wH8/Rn0sed5ajVWzQ21Y+037tO89/SP+0K735QmYpl8Fg6DJ0pAVfCrgLbfe3trm5GjgTQCn1voiEgKI2notS6i/AXwDGjh2bvvm3of2xssiHAoiTCd+gEgvYjJINKAWN+AkQYRe59KCGRuUjQITlMopjWJ3+e5Y9ql++oF6SpmLOvqVW8xPx6IIzkfr2zUrPtTwEdupHtKnzJMQZDAZDK3SkBf8RMFxEhohIAJ0091zSMRuBSQAiMgoIAWXWcVNFJCgiQ4DhwIcdOFZDK7z8/nJKfn4C42bOZsI9r/PQ0D+xU+UStkrJNiofCyIT4jXiAW71zSYgUcpULt9qmsUT0UnkUcsT0Ul8q2kWSiS9uHstl7u7AMqx3yWhtKy9b/TFOsmtvbPSfVYZ2VC+ft9ZepIbDAZDG+gwC14pFRGRHwGL0EvgHlFKrRKRWcBSpdRzwI3AX0VkBjrh7kqllAJWichT6IS8CPBDk0F/4FhYUkr9y3dzhqzW7VorriK/5l8Uep1+534i1JBBGfnNYu89PVW8HLwlwbrP8Ht59Yw3OKv0Ie1WR2krPK+/LjBTvbW5YBcMJr4Iz+t39tVs75isdL9VpCYa1hXrTJzdYDB0ITp0Hby1pv2lpG13uN5/CkxIc+4vgE5Q/eMg5+6eTI40xn09tis+Ff2ljLmBWUxunMW1vuc53/seHoF6FeCV6DjuiV6KAH3zM7jpjBGcdUw/eD5XB+HtZidDT9ECnX1mc8G2l60BjDhbZ9jXbO+4ZDd/pv4ZroWske13XYPBYNgPmEp2hpaZ/gn/vfccxnjWAVCv/OxSufSTXdb6dcEjinWqL5tVMdM8rzHN9xohGvGIbu8aJEwNGexQeXx5zzmJ10+11CyVYCdXjvv0Wf3TF+y4e7fLzIKeTBgMBkMXwgi8oRl22dktFfWMzqvnac9XgC3WkfiStojy4LWqvx3q2cKhHl3GwG3hvxU9klKK6CmV9M1P0Rq1rdb39E/gxRusmu9q/5R4dbdyzTICbzAYuhZG4A0JJJedvaT2CQLeCAisiA1hhRrC2Z4PAPhS9Wa4Zwu1KkCWNKW83iTfcgAalJ+GM0bs/cByekNWL5c7fz+UeE2w4As77nsMBoOhAzACb0jALjubnCgHcIx3PaPURi5uup1ng3dwiOjKcanEXaHz3etVgLe944l94y4mt7VGfDr2d4lXY8Ebl7PmoAAAIABJREFUDIYujBF4A5DYDQ50kZpbfbM5x/sBfokSVl78EmVS470M8OjGLnZ9+fejoxjvWe2UmPUGkGgTePxkqAhnjBkOJxy974Pc35XjPF69XC/aZGLwBoOhy2EE3tDMLQ9QRgE1ZOAjilLgRe/LlgYyaEw4v8ZfCDGxxLBRt0EF+M4cHTPvymvHfRla4I0FbzAYuhgHulStoROQrhtckVSxTvVlNzm8FNVtBB7yP0Rvdicc940hAWTsVfA/r0EgS1d+yyyEQyZpS9tteXc17LXwxoI3GAxdDCPwBzELS0qZcM/rcbe8m2LKKZBqvlB9qFKZ/CRyDQDDpJTzvO8lHlwwRAt579HQd4x1gZFJbeG6KHYcPssk2RkMhq6FcdEfpKRyy7u5zreAcbKWUgrpJ7tYFboa0J3gTvQmlZdd+nf98gWhcLje1ljdkcPff9jlao0FbzAYuhhG4A8ykpPpkknOnh8gOqEuiuhqsqKI4cVD1ElA82Xo2HukEbav1Cdu+ySxtWpXxeMFBJpqIJh9oEdjMBgMbca46Ls5tht+yMwXGXX7y1w/dzlNFVuYG5hFMRWAdsfPDcxiJBv4VA3is1jf+PkKgayeeI+cilf04jePlXBHLOLUix99MRwxxVk73l1aq9aWAQre+n8HeiQGg8GwRxgLvhuT7IavD+uqc7b7/TrffG6PXBX//EDg9wyXUrudCzHx4VERHYduqtax9lgUQrnaUk9ek57V07LoQ12/tWpyaVx3GKIreyQMBsNBg+jmbV2fsWPHqqVLlx7oYXQqkhPoUhWvaRHxanEP5cENn8Jz18Fnr8DIc+DT5+Dm9YnHz5kG2b3S15XvSlRvg0W3wZrndXc7d2ncrjppMRgM3Q4RWaaUGptqn7HguxnuOvL21K2Ycn4feCje5e1s7wcEJEpU6RiNQifPNUNFdex59Lf151AuNFRBUx0EMpsfv78L0XQkOb0hmOPySOyH0rgGg8HQjhiB70aky4y/zjefcbKWab7XqCEDvxVD91qibmu7PSGIa33xSKjYCP4s/TmYB5F6aKh0tnVn9ndpXIPBYGhHjMB3I5IL1iS75O0ubzEFUSV4UfGl6mFPCL8/AI1V2jWvrOuE6x1rPZSrf9ZsS6zT3l3pTh4Jg8Fw0GGy6LsJC0tKmy19m9h4PwsjJxJR+tfcqHwsiEzgtdgx1JKBcrnl/Xm9YcjJMPZq+P5bEMjWljoK/JbABy2Br96mK9YZDAaDodNiLPguzsKSUu58bhUV9c2T5+x68nbPdj8Rasigr+wilzoSwu7lX+mXL6it1fxBOrmueqsj5nELfjv0PrIjb8tgMBgM+4ix4Lswdsw9lbjb9JLy+Pt1qh/FUkkfbxXbep+StG49I3HdejBbu+KhuQWvYqmT7AwGg8HQaTAC34VJ1yTGzVORryECjcpLLRnclfVTBgZq6TNoRMtZ4oFsqLHWeyfH4OHgSLIzGAyGLowR+C5IuiYxxZQzP3AH8wN3xKvUfd/3AgBbVCEjZCO/q/sJNFZCdrGTJf69xfqnO0s8mKOXyIEri94l8MaCNxgMhk6NicF3MVpqEnOdbwHHyDoAPgz+IKGZ2xCPtsbHyWd6Q1bPlrPE3XXX4xZ8nrPNbwTeYDAYOjNG4LsYqdzyLVWoUyqxa2v87fPXwcs3pS+7Gshx3scteNc2k0VvMBgMnRrjou9ibEnRBW5i4/18FB2W8ngRLfIKp5ANAIdMarkRTIKYW9a61+9Y7gfDOniDwWDowhiB72L0zW8urGUU0EQASBJxCxn2DaRoZOKyuMyilsuuul30bne8HYc3LnqDwWDo1BiB7wLYSXWDZ74YT6yzW7zayXQ9pAqATbGi5hdY/2/YuQbcEl+3s+UvDbhj8C53vG3ZGxe9wWAwdGqMwHdy7KS65Ix5d8tXgDdiY2hSXr7wHQJHXgL9j3MOtte4X/qMs61gcMtf7M6Yd1vrIWPBGwwGQ1fAJNl1cn7+/KpW68tf5ltMVHnYRR4V5z0Kx/SD52dA6VLwBvQa95XzYcXTzoVb62/udtHbxXDAEX6zTM5gMBg6NcaC78QsLCmlvC4xO96uL9+k9NzMri//kedIgnk9mXxMP31g8hr3Q76uK9d5rDldcuW6ZGwXvT8TPK5/JnEL3rjoDQaDoTNjLPhOiN3TPdktD059eR8RQNeXV8EcxvcJg9/lVk+1xv35GbrMbFv6mwddAu/GniBEmo/NYDAYDJ0HY8F3Mm5buIIZc5enFHebIqlineoLwGo1iPE9IzppLjNFgp2blirXJZPOFV+2Vv8smY3BYDAYOi/Ggu9ELCwpZfaSjSmXurm5NjyDJ/y/4FC2UNDvUPp+fx78aiBktSLwe9LfPO6it1zxd/eESKOzf+U8/UoXwzcYDAbDAcVY8J2IexetTSvu7mVx9198NCcV6jrxfWPbINKk68u3ZsHvCbaL3rbgp3+iY/hevd6+1Ri+wWAwGA4oRuA7ES255e1lcTeFFjL5yF5QuVnvKP8S6nbp91mF7TcYfyYgsPNzqN4OOb31GvhYBLzB1mP4BoPBYDigGBd9J2FhSSlC80p0ycviLuJVuMsS8l5HwPaVULZGf25PC14EPH5orIK3fq1d+nYMf+x3YemjLcfwDQaDwXBAMQJ/ALGz5bdU1OMRSemen9h4P7f6ZnOO70P8RLTo+jO1S37oKVrgn7tOH9xaDL6tJMfbU62Zby2GbzAYDIYDinHRHyDcFeoUEFWJ8m7H3EGYfPxILe4AsbAWd4A+R+uflRutsxKqze89Jt5uMBgMXR5jwR8gUrV9dWPH3G/Jeg5qfZDKgT//e4mfHzu7fbLa3fH2tqyZNxgMBkOnwwj8AWBhSWnahLrkmPsF0VdgzR5cPNKoXez7KvIm3m4wGAxdGlGqtVXXXYOxY8eqpUuXHuhhtIrtmk9nvRdTzq2+2Zzr+wAfUeKWuz9LZ8lXbEx5HgDihSMugNN/Yaxtg8FgOAgQkWVKqbGp9pkY/H6mNde8XYrWi32MNQEL17rEXbSYA4Tync8qalzpBoPBYACMi36/s6WFte42g2WbfpM/MNFi73csZPTQ223X+ZoXYeyFxpVuMBgMhgSMwO9n+uZntFjQBkAhVj78/2/v3qPjKuv9j7+/mZlcml5S6M2mXLUUhSqFiggcjgcWFBVtFZCiR114QVG0iqc/QVkc5HCWaJf688JRUfF4jmhRLLH4AytXLyDYQktLi4FSLm0KbYGmNG06k8x8f3/snWaSZiYTyZ49k3xea2XN7Gf2TL7zsMs3z2U/T79Z8WMmwQd+1Xt8zjf73q6mW9dERCSkBF8mhXaIm8xOflj7LZJ0c6w9S43lzYlof7bvh2TTiIiIlEJj8GWQf897f59N3soc28hsewbDeTo3hf3zHpMN8NrTe08+59vlCVhERKqeWvBlMNDEuv63w/U4wvJub8umg/F2aoAcPPAddcOLiEhJIm3Bm9nZZtZqZhvN7PIBXv+Wma0Jf54ws/a817J5ry2PMs6oDTSxbkH6K+z1JAXvUpw5DzB4+L+BXFC26idw9YTgPncREZEiImvBm1kCuB44E9gCrDSz5e6+oeccd/983vmfAebkfUSnux8XVXzl1DQmxc69fVvrH0reSUO4/KwzwCKzT64IHq0mWDK2e1/QZf/6c4L73EVERIqIsov+RGCju28CMLOlwHxgQ4HzLwT+PcJ4YtGyuo2Ofd1MZiffq/0ux9lG6qx74JOT9UEih95kbklYd7OWjBURkSGJsou+Gdicd7wlLDuAmR0GHAHck1dcb2arzOxBM1sQXZjRaFndxinX3cPnbl5DV875XPI3vNlauS37Vlq6TybrQdV3J+qDjVy+8AS86f1Biz0/mWd2B0vGfuyu4FH3uYuISAkqZZLdQuAWd8+fiXaYu7eZ2ZHAPWa2zt2fyn+TmV0MXAxw6KGHli/aQeQvR9t/Mt15yT8DBGPvliCZy/S2ygda/33hTb0frAl2IiJSoigTfBtwSN7xjLBsIAuBT+cXuHtb+LjJzO4jGJ9/qt85NwA3QLAW/bBEPQzyZ8337Of+rsRfSZjT6bXs9VoOrumAUz8Hne29rXIlcxERGSZRJviVwEwzO4IgsS8E3t//JDM7GpgI/DWvbCKw193TZjYJOAX4eoSxDqv8WfM9a8vXhGvK19FFB/XBi7tfgAX/FUeIIiIywpU8Bm9mY4bywe7eDVwKrAAeB37l7uvN7Boze3feqQuBpd53W7vXA6vM7FHgXuC6/Nn3lW56U0Of40n2CnupBaDGnMk1rwQvrLlJt72JiEgkBt0u1sxOBn4MjHX3Q83sTcAn3P1T5QiwVJW0XeyBW8I6T9R9iFrLsjJ3NHNr/h7cFpd/25tmxouIyBC92u1ivwXMA14CcPdHgdOGL7yRZ8GcZv5zwTH7jyfYPmotSPZHTcgFyb0mpdveREQkMiWNwbv7ZrM+S7EU3tBcAHjjIRMB+Mb5b+Lcw9Pw3aB8Qnpr8GT+f8HmB3Xbm4iIRKKUBL857KZ3M0sBiwjG1KWAltVtfOW29QBc9/u/M+mkNP/c82KmI3g8+u3wpvfFEp+IiIx8pXTRf5LgFrZmgtnwx9Hvljbp1TP+nty7nZtrr4Hd21h635rgxVQ4T3H8DKgbF1+QIiIy4hVtwYfryX/b3T9QpniqXs898F9K3sqbrZXPJpexNnckJIBps2HzQzB5VtxhiojICFc0wbt71swOM7Nad8+UK6hq1LK6jSUrWrm7833U1/euXPfB5F29Jx30uiDBTzhkgE8QEREZPqWMwW8C7g+3bN3TU+juWmotlH9b3D8RrFz37sQD1Bh0ei1bfBKH1uyg7uVNwRteejLegEVEZMQrJcE/Ff7UABo4HkD+0rQ7mEgntRjBevN1dPFa2xpMdtgcLtb37P3BAjfJOrhye1xhi4jICDZognf3rwCY2djwuCPqoKpN/tK0AIfbNszgFa+nJXsq70w+xMH1CehOQ3en9nUXEZHIDTqL3syONbPVwHpgvZk9bGbHDPa+0aT/0rS35U4GYAwZruq+iC25yZBqDBa20b7uIiJSBqXcJncDcJm7H+buhwFfAH4UbVjVZfG8WTSkEkxmJzfXXsPJtg6ApOUYQ5qpNbtg7w5440Lt6y4iImVRyhh8o7vf23Pg7veZWWOEMVWdBXOa6c7m6GxZxJutlXSit1onp9JMoR2yWUg1BLfKaStYERGJWCmbzdwKPAL8b1j0r8AJ7v6eiGMbklg3m7l2SjC+XipNrhMRkWHwajeb+QgwGVgG/AaYFJZJj0Vref6Qc+jyAapzzMG9z5MNMPt8WLSufLGJiMioVMos+p3AZ8sQS9Vq2ZglsznN+eTo6Q8xqwHPAeEmPdo9TkREyqiUWfR3mllT3vFEM1sRbVjVo2eRm/HZdp73iez2Bp70Gbw8IbzRoGdzmXd9W5PrRESkbErpop/k7u09B2GLfkp0IVWXr9y2ns6uLJ/s+jx7GMP9uWM5K/11LupcFJww/fjg8ZC3BJPrFt4UX7AiIjJqlJLgc2Z2aM+BmR0GFJ+ZN0q0rG5j597edeen2k5e8IMA2LgrrNqe5WkbD+7/dhERkciUcpvcl4G/mNkfCQaU/wm4ONKoqsSSFa37nzewj/G2l+0+EYCmCU2QSUDHC1CThPqmQh8jIiIy7EqZZPd7MzseOCks+py7vxhtWNUhf4naKRaMYmzzIJEvPvto+MN46NwZzKQ3iyVGEREZnUqZZHcK0OnuvwOagC+F3fSjXv4StVPZCcA2JtLUkGLBnOZgxjz0vVVORESkDEoZg/8+sNfM3gRcRrCz3P9EGlWVWDxvFskaYzI7+c/UTwDYlZjE1e8OZ9DXK8GLiEg8Sknw3R4sdzcfuN7dr0fbxsLuF1iw5mOccFCGL6Zu5rW2FYBL3nVK0HoHqJsQPDZOiilIEREZrUqZZLfbzK4gWKL2NDOrAVLRhlUF7vsa/uwD3MwDkOgtfuftJ8EfwqVo97fgleBFRKS8SknwFwDvBz7q7i+Et8wtiTasCpa37vyA0+YOOhIu+n3wvGcMXi14EREps0G76N39BXf/prv/OTx+zt1H7xj8orVsbn4nPXv0HLBXz8ub4BtHBX8I1Idd9BqDFxGRMitlDF7yjZvGo8939rnrrdutd6MZq+ndUMbCskQpHSUiIiLDRwn+H9DU1buevBl0UkcCZ5+HUxN6NpRpC7evfUJL94uISHmVch/8u8KJdRK6pf68PsdtPomfZ8/gE3VfCzaUefincPUE2LIyOKH1juD4Wi3hLyIi5VFK4r4AeNLMvm5mR0cdUDX44LF1AOzxWgCeyk3nq/Zx3vP2s4MNZS57HI49H5L1wRu0D7yIiJRZKZPs/hWYQ7DAzX+b2V/N7GIzG7X3wp/w6FUANFoGgHcm/8bjiQtY8P/mBCeMmwZ14yCbCZK89oEXEZEyK6nr3d1fAW4BlgKvAd4DPGJmn4kwtso1+wJyNSk6wxb8gC30PduD7vqP3aV94EVEpOwGnd5tZu8GLgJeR7BE7Ynuvt3MxgAbgO9GG2IFSu+iO9FAXXY3uUQdNQO10PP3fT/nm+WPUURERrVS7t86F/iWu/8pv9Dd95rZR6MJq8Ltfp5uEizNnsGCi77M+Md+rha6iIhUlFK66K8G/tZzYGYNZnY4gLvfHUlUFW7vS22syLyRq7o/wtuXttPS/IW+LXYREZGYlZLgfw3k8o6zYdmo1PLIZmo7t3GiP8Zk2mlr7+SKZetoWd0Wd2giIiL7lZLgk+6e6TkIn9dGF1Jl+/GKlSTNeY29xGeTywDo7MqyZEVrzJGJiIj0KmUMfoeZvdvdlwOY2XzgxWjDqlDXTuF34UYzNQYfTN7FB5N3sc9TvL79ZzEHJyIi0quUFvwngS+Z2XNmthn4IvCJaMOqUIvW8ogds/+w02u5tfsU/in9baY3NcQYmIiISF+DtuDd/SngJDMbGx53RB5VpRo3jYljkrAH0p6kji46aKAjdTBfnTcr7uhERET2K2mbMzN7J3AMUG/hNmrufk2EcVWkltVtzO7YgQPvzVzNBYn7mGK7OPeEZhbMaY47PBERkf1KWejmB8AY4F+AHwPnkXfb3GiyZEUrl2ZnMS6xl/V+JFd1HwlA8993xByZiIhIX6WMwZ/s7h8Cdrr7V4C3AkdFG1Zl2treyVTbyTZvOqBcRESkkpSS4PeFj3vNbDrQRbAe/agzvamBabaTbT7xgHIREZFKUkqCv83MmoAlwCPAM8AvogyqUi2eNytswR+0v6whlWCxJtiJiEiFKToGb2Y1wN3u3g78xsx+B9S7+66yRFdhFsyeDL99he1MxAha7ovnzdIEOxERqThFE7y758zseoL94HH3NJAu9cPN7Gzg20AC+LG7X9fv9W8RTN6DYCLfFPdggNvMPgxcGb52rbvHv5JMuKFMqmk6T//bO2MORkREpLBSbpO728zOBZa5u5f6wWaWAK4HzgS2ACvNbLm7b+g5x90/n3f+Zwj/kDCzg4B/B+YCDjwcvndnqb8/ErufB8DHjsopCCIiUkVKGYP/BMHmMmkze8XMdpvZKyW870Rgo7tvCtevXwrML3L+hcAvw+fzgDvd/eUwqd8JnF3C74zW9scBGDd2TMyBiIiIFDdognf3ce5e4+617j4+PB5fwmc3A5vzjreEZQcws8OAI4B7hvJeM7vYzFaZ2aodO6K/F737kWBL2JM67or8d4mIiLwapSx0c9pA5e7+p2GMYyFwi7tnh/Imd78BuAFg7ty5JQ8fDNm1U6A7vb+yjn7hNrh6AiTr4Mrtkf1aERGRf1QpY/CL857XE3S9PwycPsj72oBD8o5nhGUDWQh8ut9739bvvfcNHmpEFq1l89LLmN52BwlydFLLizPO4pALvhlbSCIiIsWU0kX/rryfM4FjgVImu60EZprZEWZWS5DEl/c/ycyOBiYCf80rXgGcZWYTzWwicFZYFouWjVn+sjlNjefIulHnXfzluTQtG7vjCklERKSoUibZ9bcFeP1gJ7l7N3ApQWJ+HPiVu683s2vM7N15py4ElubP0Hf3l4H/IPgjYSVwTVgWiyUrWpnou0iT5I7sm/l59gwmejtLVrTGFZKIiEhRpYzBf5fgVjUI/iA4jmBFu0G5++3A7f3Krup3fHWB994I3FjK74na1vZO/g8fZ239xazxmfw4G9wDb1qDXkREKlQpY/Cr8p53A7909/sjiqciTW9q4KBdmwB4zqf0KRcREalEpST4W4B9PTPczSxhZmPcfW+0oVWOxfNm8adb7wPgWZ8KaA16ERGpbKWMwd8N5DdVG4BRdSP4gjnNfOyY4PlzPoXmpga++t7ZWoNeREQqVikt+Hp37+g5cPcOMxt1S7nN9OfIeIIrTp/Bh856S9zhiIiIFFVKC36PmR3fc2BmJwCjb3bZc/eTIsupWyti3p+IiEhRpbTgPwf82sy2AgZMAy6INKpKEq5ilwIwOPKZpXD1Uq1iJyIiFa2UhW5WAkcDlwCfBF7v7g9HHVjFWLQWjj1v/32CuWQDzD4fFq2LNSwREZFiBk3wZvZpoNHdH3P3x4CxZvap6EOrEOOmQaoeA7q8BsumoW48jJsad2QiIiIFlTIG/3F3b+85CLdv/Xh0IVWgV4J94H+aOBc74SLo2BZzQCIiIsWVkuATZmY9B2aWAGqjC6ny3HNIsA/Ow/uaOeWxc2iZ9fWYIxIRESmulAT/e+BmMzvDzM4AfhmWjQotq9v42d2rAdjFWNraO7li2TpaVhfaGE9ERCR+pST4LwL3EEyyu4Rg4ZvFRd8xgixZ0Up9djcAu7wRgM6urDaaERGRilbKLPqcu//A3c9z9/OADcB3ow+tMmxt76TJ9gC9Cb6nXEREpFKVch88ZjYHuBB4H/A0sCzKoCrJ9KYGJuwOFvJrZ2yfchERkUpVMMGb2VEESf1C4EXgZsDc/V/KFFtFWDxvFttu3UuXJ9hLHaCNZkREpPIVa8H/HfgzcI67bwQws8+XJaoKsmBOM4+vTNLe1ggYzU0NLJ43SxvNiIhIRSuW4N8LLATuNbPfA0sJlqoddabXp3nRG/nhB09g3jHT4g5HRERkUAUn2bl7i7svJFim9l6CNemnmNn3zeyscgVYCbyznV00Mq6upCkLIiIisStlFv0ed/+Fu78LmAGsJrh1btSo2ddOu4+lUQleRESqRCn3we/n7jvd/QZ3PyOqgCpRIr2LXTQqwYuISNUYUoIfrZJdu9jljYxVghcRkSqhBD+YXI7art3sYiyNdYm4oxERESmJEvxg0rswnF3eSGOtWvAiIlIdlOAH8+ITANQkEtTUjMq7BEVEpAopwQ/mge8B8LaaR2MOREREpHTqcy7k2inQnd5/eCqr4eoJkKyDK7fHGJiIiMjg1IIvZNFaOPZ8SNQCkCEFs8+HRetiDkxERGRwSvCFjJsGdeMg2wVAkm6oGw/jpsYcmIiIyOCU4IvZsx0OPw2AhxpPh45tMQckIiJSGiX4YhbeBMe+B4A7pn0yOBYREakCSvCDyWYAqK1viDkQERGR0inBDyacSV+vBC8iIlVECX4Q2a59ANTVjYk5EhERkdIpwQ+iKxMk+DEN9TFHIiIiUjotdDOIrvQ+zFOMrVdViYhI9VALvoiW1W38bvUzpEly3R1/p2V1W9whiYiIlEQJvoCW1W1csWwd2cw+MqTYubeLK5atU5IXEZGqoARfwJIVrXR2Zamlm0w4ktHZlWXJitaYIxMRERmcEnwBW9s7Aai1LjKeOqBcRESkkinBFzC9KbjvPb8Fn18uIiJSyZTgC1g8bxYNqQS1dAU7yQENqQSL582KOTIREZHB6d6vAhbMaQag8bdZMrkk08bXc/nbj95fLiIiUsmU4ItYMKeZ5/+Y5NmXU/zhstMYX58a/E0iIiIVQF30g7BshownqU8m4g5FRESkZErwg6jJZciQIpWwuEMREREpmRL8IGqyGbqsFjMleBERqR6RJngzO9vMWs1so5ldXuCc95nZBjNbb2a/yCvPmtma8Gd5lHEWU5PLkK3R2LuIiFSXyCbZmVkCuB44E9gCrDSz5e6+Ie+cmcAVwCnuvtPMpuR9RKe7HxdVfKVK5LrI1tTGHYaIiMiQRNmCPxHY6O6b3D0DLAXm9zvn48D17r4TwN23RxjPPyThGXJK8CIiUmWiTPDNwOa84y1hWb6jgKPM7H4ze9DMzs57rd7MVoXlCwb6BWZ2cXjOqh07dgxv9KFkTgleRESqT9z3wSeBmcDbgBnAn8xstru3A4e5e5uZHQncY2br3P2p/De7+w3ADQBz5871SAL0LiV4ERGpOlG24NuAQ/KOZ4Rl+bYAy929y92fBp4gSPi4e1v4uAm4D5gTYawDy+VI0o0nlOBFRKS6RJngVwIzzewIM6sFFgL9Z8O3ELTeMbNJBF32m8xsopnV5ZWfAmyg3LIZACV4ERGpOpF10bt7t5ldCqwAEsCN7r7ezK4BVrn78vC1s8xsA5AFFrv7S2Z2MvBDM8sR/BFyXf7s+7LJpoPHZF3Zf7WIiMirEekYvLvfDtzer+yqvOcOXBb+5J/zADA7ythK0h204FELXkREqoxWsiumpwWfUAteRESqixJ8Md1BgreUEryIiFQXJfhiwkl2pjF4ERGpMkrwxfS04JXgRUSkyijBF+Fhgq9RF72IiFQZJfgiujL7ACV4ERGpPkrwRXSlgwSfTNXHHImIiMjQKMEX0ZXpBCBRqwQvIiLVRQm+iK50MAafUAteRESqjBJ8EdmuoAWfqtMYvIiIVBcl+CK6M0ELPplqiDkSERGRoVGCLyLbFUyyS9Wpi15ERKqLEnwR2bAFn6pTC15ERKqLEnwRuZ4xeM2iFxGRKqMEX0TPSnZ19WrBi4hIdVGCLyLXlabLE9SlknGHIiIiMiRK8EV4d5oMSepTibiGa8QLAAAMO0lEQVRDERERGRIl+CIS6XZq6aJ+3464QxERERkSJfgiDt61jiQ5Gh/6RtyhiIiIDIkGlwdy7RToTnMQgEHqkZ/CIz+FZB1cuT3u6ERERAalFvxAFq2FY88n11M9yQaYfT4sWhdvXCIiIiVSgh/IuGk83VGDeY6cG7nufWzanYBxU+OOTEREpCTqoh9Ay+o2xmzaRKNNoC03iXV+BNM2PcXa1W0smNMcd3giIiKDUoIfwJIVrbRlPs8dtZezgyau6v4IAM0rWpXgRUSkKqiLfgBb24MlahvpZA/1B5SLiIhUOiX4AUxvCpambbR97PH6A8pFREQqnRL8ABbPm0VDKkEj++ggSOoNqQSL582KOTIREZHSaAx+AAvmNGO5bupv62Kv19Pc1MDiebM0/i4iIlVDCb6A+W8YD7fB5EkHc//nT487HBERkSFRF30hmT0AdCfGxByIiIjI0CnBF5LuAKA7qQQvIiLVRwm+kLAFn02NjTkQERGRoVOCLySzG4BcqjHmQERERIZOCb6QsIteCV5ERKqREnwhYRd9Tl30IiJShZTgCwm76KlVghcRkeqjBF+Ah1301KmLXkREqo8Wuikgt68Dc6OmVgleRESqjxJ8Adn0btLUU5tKxB2KiIjIkKmLvoDcvt3spY7ahKpIRESqj7JXAZ7uoMMbqE2qBS8iItVHCb6QdEfQgk+qikREpPooexWS6WAPDUrwIiJSlZS9CsnsocPrqVOCFxGRKqTsVYB1dbCXerXgRUSkKil7FVCTfoW32AbGZl6KOxQREZEhizTBm9nZZtZqZhvN7PIC57zPzDaY2Xoz+0Ve+YfN7Mnw58NRxjmQZGYXU2wXhz32vXL/ahERkVfN3D2aDzZLAE8AZwJbgJXAhe6+Ie+cmcCvgNPdfaeZTXH37WZ2ELAKmAs48DBwgrvvLPT75s6d66tWrXr1gV87BbrTB5Yn6+DK7a/+80VERIaJmT3s7nMHei3KFvyJwEZ33+TuGWApML/fOR8Hru9J3O7ek0HnAXe6+8vha3cCZ0cYa69Fa+HYc8lZsMhfLlEPs8+HRevK8utFRESGQ5QJvhnYnHe8JSzLdxRwlJndb2YPmtnZQ3gvZnaxma0ys1U7duwYnqjHTYO6CZhn2ecpLJuGuvEwburwfL6IiEgZxD3JLgnMBN4GXAj8yMyaSn2zu9/g7nPdfe7kyZOHL6o929l42Pt4T+Ya9rzxw9Cxbfg+W0REpAyi3GymDTgk73hGWJZvC/CQu3cBT5vZEwQJv40g6ee/977IIu1v4U08+NdneLx1PZ1nfpSx4+rK9qtFRESGQ5Qt+JXATDM7wsxqgYXA8n7ntBAmcjObRNBlvwlYAZxlZhPNbCJwVlhWNunuHIDugxcRkaoUWQve3bvN7FKCxJwAbnT39WZ2DbDK3ZfTm8g3AFlgsbu/BGBm/0HwRwLANe7+clSxDiSTDRK8VrITEZFqFOl+8O5+O3B7v7Kr8p47cFn40/+9NwI3RhlfMZmeFry2ixURkSqk7FVAujtHssaoqbG4QxERERkyJfgCMt05dc+LiEjVUgYrINOd0wQ7ERGpWspgBSjBi4hINVMGKyCTVYIXEZHqpQxWQKY7pxn0IiJStZTBCkh3Z6lNJuIOQ0RE5B+iBF9AWrPoRUSkiimDFaBJdiIiUs2UwQrIZNWCFxGR6qUMVoAm2YmISDVTBitAXfQiIlLNlMEKSCvBi4hIFVMGK0Br0YuISDVTBitAK9mJiEg1UwYbQMvqNnbuyfDzB5/jlOvuoWV1W9whiYiIDIkSfD8tq9u4Ytk6PDxua+/kimXrlORFRKSqKMH3s2RFK51d2T5lnV1ZlqxojSkiERGRoVOC72dre+eQykVERCqREnw/05sahlQuIiJSiZTg+1k8bxYNqb67yDWkEiyeNyumiERERIYuGXcAlWbBnGYgGIvf2t7J9KYGFs+btb9cRESkGijBD2DBnGYldBERqWrqohcRERmBlOBFRERGICV4ERGREUgJXkREZARSghcRERmBlOBFRERGICV4ERGREUgJXkREZARSghcRERmBlOBFRERGIHP3uGMYFma2A3h2mD92EvDiMH9mNVN99FJd9KX66Ev10Zfqo6/hrI/D3H3yQC+MmAQfBTNb5e5z446jUqg+eqku+lJ99KX66Ev10Ve56kNd9CIiIiOQEryIiMgIpARf3A1xB1BhVB+9VBd9qT76Un30pfroqyz1oTF4ERGREUgteBERkRFICX4AZna2mbWa2UYzuzzueOJgZs+Y2TozW2Nmq8Kyg8zsTjN7MnycGHecUTGzG81su5k9llc24Pe3wHfC62WtmR0fX+TRKFAfV5tZW3iNrDGzd+S9dkVYH61mNi+eqKNjZoeY2b1mtsHM1pvZorB81F0jRepiVF4fZlZvZn8zs0fD+vhKWH6EmT0Ufu+bzaw2LK8LjzeGrx8+bMG4u37yfoAE8BRwJFALPAq8Ie64YqiHZ4BJ/cq+DlwePr8c+FrccUb4/U8DjgceG+z7A+8A7gAMOAl4KO74y1QfVwP/NsC5bwj/3dQBR4T/nhJxf4dhro/XAMeHz8cBT4Tfe9RdI0XqYlReH+F/47Hh8xTwUPjf/FfAwrD8B8Al4fNPAT8Iny8Ebh6uWNSCP9CJwEZ33+TuGWApMD/mmCrFfOBn4fOfAQtijCVS7v4n4OV+xYW+/3zgfzzwINBkZq8pT6TlUaA+CpkPLHX3tLs/DWwk+Hc1Yrj78+7+SPh8N/A40MwovEaK1EUhI/r6CP8bd4SHqfDHgdOBW8Ly/tdGzzVzC3CGmdlwxKIEf6BmYHPe8RaKX6wjlQN/MLOHzezisGyquz8fPn8BmBpPaLEp9P1H8zVzadjlfGPekM2oqo+wS3UOQUttVF8j/eoCRun1YWYJM1sDbAfuJOilaHf37vCU/O+8vz7C13cBBw9HHErwUsip7n488Hbg02Z2Wv6LHvQnjdpbMEb79w99H3gtcBzwPPCNeMMpPzMbC/wG+Jy7v5L/2mi7Rgaoi1F7fbh71t2PA2YQ9E4cHUccSvAHagMOyTueEZaNKu7eFj5uB24luEi39XQrho/b44swFoW+/6i8Ztx9W/g/shzwI3q7WUdFfZhZiiCh3eTuy8LiUXmNDFQXo/36AHD3duBe4K0EwzLJ8KX877y/PsLXJwAvDcfvV4I/0EpgZjjjsZZg0sPymGMqKzNrNLNxPc+Bs4DHCOrhw+FpHwZ+G0+EsSn0/ZcDHwpnSp8E7Mrrph2x+o0hv4fgGoGgPhaGs4OPAGYCfyt3fFEKx0h/Ajzu7t/Me2nUXSOF6mK0Xh9mNtnMmsLnDcCZBPMS7gXOC0/rf230XDPnAfeEvT+vXtwzDivxh2DG6xME4yZfjjueGL7/kQSzXB8F1vfUAcG40N3Ak8BdwEFxxxphHfySoFuxi2C87KOFvj/BrNnrw+tlHTA37vjLVB//G37fteH/pF6Td/6Xw/poBd4ed/wR1MepBN3va4E14c87RuM1UqQuRuX1AbwRWB1+78eAq8LyIwn+kNkI/BqoC8vrw+ON4etHDlcsWslORERkBFIXvYiIyAikBC8iIjICKcGLiIiMQErwIiIiI5ASvIiIyAikBC8iAJhZNm/nrzU2jDspmtnh+TvRiUj0koOfIiKjRKcHy2uKyAigFryIFGVmz5jZ181sXbjP9evC8sPN7J5wM5G7zezQsHyqmd0a7of9qJmdHH5Uwsx+FO6R/YdwlS8RiYgSvIj0aOjXRX9B3mu73H028D3g/4Zl3wV+5u5vBG4CvhOWfwf4o7u/iWAP+fVh+Uzgenc/BmgHzo34+4iMalrJTkQAMLMOdx87QPkzwOnuvincVOQFdz/YzF4kWH60Kyx/3t0nmdkOYIa7p/M+43DgTnefGR5/EUi5+7XRfzOR0UkteBEphRd4PhTpvOdZNAdIJFJK8CJSigvyHv8aPn+AYLdFgA8Afw6f3w1cAmBmCTObUK4gRaSX/oIWkR4NZrYm7/j37t5zq9xEM1tL0Aq/MCz7DPBTM1sM7AAuCssXATeY2UcJWuqXEOxEJyJlpDF4ESkqHIOf6+4vxh2LiJROXfQiIiIjkFrwIiIiI5Ba8CIiIiOQEryIiMgIpAQvIiIyAinBi4iIjEBK8CIiIiOQEryIiMgI9P8B7qmP3p0lWuUAAAAASUVORK5CYII=\n",
            "text/plain": [
              "<Figure size 576x432 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "tUc5ZxdLVaXc",
        "outputId": "45fc6e35-f6c8-49b0-8673-fd79e05bfd50",
        "colab": {
          "base_uri": "https://localhost:8080/"
        }
      },
      "source": [
        "test_loss,test_acc = model_NN.evaluate(X_test_scaled, Y_test, verbose=1)\n",
        "print('Test loss:', test_loss)\n",
        "print('Test ACC:', test_acc)"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "161/161 [==============================] - 0s 980us/step - loss: 0.3939 - accuracy: 0.8340\n",
            "Test loss: 0.3938627243041992\n",
            "Test ACC: 0.8340458869934082\n"
          ],
          "name": "stdout"
        }
      ]
    }
  ]
}