This code was written by Mu Tian (kevinmtian@gmail.com) and Zhonglin Han (hanzhonglin1990@gmail.com). It provides our best submission to the Kaggle Santander Product Recommendation Competition. We ranked 16 (12 in public leaderboard). To view a write up for our approaches, please check Kaggle Santander Product Recommendation Solution(Public LB 12nd) at Zhonglin Han's Blog.
[python] Recommender.py
Configs (parameters, feature selection, data paths) for our model can be specified in helper.config
This code requires Python, numpy/scipy, scikit-learn, LightGBM, XGBoost and pandas