The Radicchio Triple Scorer
Enviroment:
Python 2.7, Scikit-Learn 0.18, Numpy 1.11.1, Scipy 0.18.1
Execution:
Change your current path to the src/ folder, and type in "python *.py".
You may first try with "python main.py".
There are two frameworks - supervised framework and unsupervised framework. You can inherits the abstract class SupervisedBase/UnsupervisedBase and implements different methods about feature extraction/learning/prediction to start your work. For details, please refer to the code and the comments.
NOTE: The data is too large, so you may download it by yourself. You should have all the data files stored in the data/ folder (try ln -s !!).
Train the models, please first run "main.py".