update model if false positive

[Pikachuzzz]

Hello,
In the deeplog document, it is said that when false positives are found, we must update the workflow.
I don't feel like it's working, am I wrong or not?
P.S.: Thank you for this complexe algorithm.

[amineebenamor]

Hello @Pikachuzzz ,
This part of the paper isn't implemented in the code of this repository. It needs to be done somehow.
If you find how to do it, do not hesitate to share your solution here.

[hayhan]

It looks the DeepLog author meant the incremental learning for the model online update. Then I have the concern how DeepLog handles the catastrophic forgetting issue that general incremental learning faces.