Anomaly detection in machine learning

Hello guys,

I am extremely interested in anomaly/fraud detection in machine learning. I have read some scientific papers about this topic and personally think that this topic is quite satured by scientific research. Besides that I would like to contribute to this field in my free time and I am looking for research questions.

Therefore, my question is:

Which questions are still open in anomaly/fraud detection? What are interesting subfield in anomaly/fraud detection?

I really would appreciate your hints about papers/research articles?

Best regards!