/MLog

Log anomaly detection based on MogrifierLSTM + CNN

Primary LanguagePython

MLog

Log anomaly detection based on MogrifierLSTM + CNN MLog uses an innovative feature extraction method combining Bert pre training model and Event_IDF. The training model adopts the combination of MogrifilerLSTM + CNN.

MLog uses "CUDA" to train and predict under windows. The required Python package is in requirements.txt. Here, we would like to thank the following works:

logpai:https://github.com/logpai

logdeep:https://github.com/donglee-afar/logdeep