parameters anomaly detection model
Closed this issue · 4 comments
hey, i tried to understand your repo many time but did not understand pretty well even from your documentation as you are not taking directly parsed input do some kind of pre-processing.
have you worked on parameters anomaly detection models and if workflow model?
I am sorry, but I am not sure if I understand your question.
This implementation can parse security events given in .txt or .csv format as specified here: https://deeplog.readthedocs.io/en/latest/reference/preprocessor.html#formats
If you have logs in a different format, you will have convert the logs yourself into one of the supported formats.
For the evaluation on the HDFS dataset, we used the already converted logs in .txt form found here: https://github.com/wuyifan18/DeepLog/tree/master/data
what about prameter anomaly detection did you implemented that part ?
No, this library only implements the Log Key Anomaly Detection model. I will probably not add the Parameter Value Anomaly Detection model myself, but feel free to send me a pull request if you decide to implement it yourself and want it added to the library.
I had rough code once optimize it will share it.