dotnet / dotnet/machinelearning
Anomaly detection evaluator throws
Open
area-Core
Priority:2
- Dominant language
- C#
- Stars
- 9.4k
- Forks
- 2k
- Avg merge
- 2d 20h
- Merged PRs (30d)
- 11
Description
Issue number #2644 was fixed in pr #2692, but now there is an exception thrown when the dataset contains only positive examples.
In my opinion, it makes more sense to report an AUC of NaN than to throw - the other metrics can still be computed, so there is no need to throw.
Contributor guide
Assessment
This issue has not been assessed yet.