- Archit Sachdeva
- Georgi Georgiev
- Krzysztof Wroblewski
classifier/
constants.py
-- use to set the dataset pathlogistic_regression.py
-- our implementation of a binary logistic regression classifiermetrics.py
-- our implementation of metrics used during training and validationmord.py
-- trains and evaluates an ordinal regression model; requires themord
packagemulticlass.py
-- our implementation of one-vs-rest classification; contains a wrapper that can be used with any binary classifierpredict.py
-- uses a previously trained model to generate a ranking file, or simply evaluate on a test setreader.py
-- contains code for reading and parsing dataset filestrain.py
-- trains the logistic regression model
eval.py
-- contains evaluation code and our implementation of ERR and NDCG metrics