JDATA-2018 用户购买时间预测

my solution for JDATA 2018
rank: A:171/5182 B:30

How to run

• extract feature from data
  --extract_fea.py
• train xgboost
  -- model_xgb.py
• ensemble mutiple models using stacking method
  -- model_stacking.py
• ensemble the results of xgboost and stacking through weighted average
  -- model_weighted.py