JData-2018

The code for JData-2018

Rank: 6th

How to Run

run the following file to create necessary features dataset

-- Feature_Engineering.sh
run one of the models to generate good S1 result with different constructed data.

-- Run_11.py

-- Run_12.py

-- Run_21.py
In addition, feature selection is one of th key process for performance improvement. To run the feature selection code, please refer to the file

-- Select.py
or the selection file for general dataset on both S1 and S2

-- Select_com.py

-- The log file is the output from t he features selection process
The details of the selection library usage can refer to my another repo