Filter noise feedback, text classification combining LSTM, TFIDF stacking, XGBoost, word2vec, LDA, LSI...

data_preprocess.py:
--Preprocess data
tfidf_stacking.py:
--Generate traditional models (lr, svc, mnb, bnb) stacking features based on Tfidf vectors
semantic_features.py:
--Generate semantic features based on Word2vec, Doc2vec, LDA, LSI
my_lstm.py:
--Generate LSTM stacking features

my_xgboost.py:
--Train XGBoost model based on Layer1 features
excute.py:
--Excute project