/JDATA-2018

JDATA 2018用户购买时间预测

Primary LanguagePython

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