The directory contains the main python files used for preprocessing and data aggregation as well as for feature training.
Here follows a brief description of the files in the current folder:
- Preprocessing.py: Prepares data and generates training-test splits 
- Station_profile_all.py: Implements training of the StationProfileAll feature
- Station_profile.py: Implements training of the StationProfile feature
- User_profile.py: Implements training of the UserProfile feature
- User_specific_model: Support class for the UserProfile feature

To run the code, data should be located in directory “../data” and the following actions must be taken:
1) Run the data preparation script with “python preprocessing.py”
2) Run the script corresponding to the feature one wants to train, e.g. for UserProfile run “python user_profile.py”

The trained model is saved by pickle in “../data/models”.