Model consisting in the combination of a neural network, a support vector classifier and a adaboosted random forest.
To test our models, models and predictions are in the non _test folders and files. The files contained in these folders shouldn't be changed. They were used for the final prediction.
To generate code please save models and csv files in folders or csv files ending by _test.
Launch files in these order without forgeting to uncomment some execution lines in files:
- feature_creation_and_preprocessing.py
- feature_selection.py
- train models and predict: neural_network.py, support_vector_classifier.py and adaboost_random_forest.py
- vote_from_csvs.py