folder 'analyse':
    some details about every attributes,'columnsType.txt' includes every attributes' different value-domain and their value-type without any preprocessed;
    and 'columnstype2.txt' includes the same content as 'columnsType.txt',but do some preprocessed operators for missing-value and then encode 'object' type attributes;

folder 'code':
    'code.py' :use to do some preprocessing operators;
    'learning.py':use to do some analysing operatores;
    'model.py':use to build some models,such as xgboost;
folder 'data':
    '[new] yancheng_train_20171226.csv':the training data download from official website;
    'yancheng_testA_20171225.csv':the test data download from official website;
    folder 'preprocessed':include data file which has been preprocessed;
folder 'plot':
    including some plots;
folder 'result':
    use to store the result