Code for Coupon Purchase Prediction (RECRUIT Challenge).
Note: This code is able to achieve a 5th place score (Private LB: 0.008776). But this is not a full version of my submitted solution (Private LB: 0.008905). My submitted solution is average of this solution and another XGBoost solution. This repositoy provides a simple version of 5th place solution.
- OS: Ubuntu 14.04
- Python: 2.7
- pip: numpy(1.9), scipy, pandas, sklearn, sklearn-pandas, chainer
Place the data files into a subfolder ./data. And unzip. (requires coupon_list_train.csv
, coupon_list_test.csv
, user_list.csv
and coupon_detail_train.csv
)
validation-set: last 4 weeks.
$ python make_data.py --validation
$ python train.py --validation
$ ./run.sh
$ ls -la submission_mlp.csv
It takes around 5 hours on 4 core CPU and 16GB RAM.
If you get an out-of-memory error, use run_low_memory.sh
.