Python scipts and notebooks playing around with the Kaggle Titanic dataset .
The main focus so far is on a Neural Network binary classifier (even though this might not be the optimal classifier for this type of data),
as a playground for building NN classifiers using Keras and hyperparameter tuning.
- Some exploration and feature engineering in notebooks/explore.ipynb
- KFold cross validaion in scripts/NN_classifier_CV.py
- Actual training on full training data and prediction on test data in scripts/NN_classifier.py
- Exploration of other classifiers in notebooks/Multiple_classifiers.ipynb (work in progress)