python 3.6
numpy
pandas
sklearn
tensorflow

1. first run preprocess.py. 
    first edit the preprocess.py file
    Two important parameters 'mainTableFile' and 'codeStateFile' should be assigned according to those two files' path
    Left 'embeddingSize' to the default value 8 
    After the edition, just use 'python preprocess.py' to run the code
2. Then run makePredictions.py to generate prediction on test files. This will generate the 'cv_predict.csv' file recording all the predctions
    edit the file , just to assign the CV folder's path 
    Using 'python makePredictions.py' to run the code
3. The last step is to run evaluation.py to caculate indices. 
    just run the file using similar commond
    
This would generate Three required file 'cv_predict.csv' , 'evaluation_by_problem' and 'evaluation_overall'