The purpose of this repo is to provide a template of anyone wishing to evaluate contextual bandit systems. It provides three functions: Data Generation, Training, and Evaluation.
Training
The part of the repo can be found at taming_the_monster/train. The modules used here are designed to assist with training a contextual bandit system. The modules that I recommend you customize are data_iterator.py and model.py. Descriptions of what each module does can be found in the docstrings.
Evaluation
The part of the repo can be found at taming_the_monster/evaluate. The modules used here are designed to assist with evaluating a contextual bandit system. The modules that I recommend you customize are model_evaluation.py. Descriptions of what each module does can be found in the docstrings.