Read me

The test_setup.py is the file where I did my preliminary analysis, but I commented it out.
The feature_extraction.py just run it after changing the address of training data and testing data and sample data. The work I've tried is commented out and the remaining is what I used for submission.
The analysis.py is the file I used to train the models and make predictions, just run the file it will work. And all tries of other models and operations which is not used are commented.
I did not upload dataset, as github didn't allow me to do that, it says the dataset is too big. visualization and analysis are commented as the work only shows the final result. If visualization and analysis work needed, it can be uncommented.