dotnet / dotnet/machinelearning
Multiclass Classification catalog does not contain EvaluateNonCalibrated
Open
area-Core
bug
Priority:2
- Dominant language
- C#
- Stars
- 9.4k
- Forks
- 2k
- Avg merge
- 2d 20h
- Merged PRs (30d)
- 11
Description
The Multiclass Classification catalog has non calibrated trainers (`SdcaNonCalibratedMulticlassTrainer`) along with many calibrated trainers.
Currently there is only one `Evaluate` method, which assumes that the model is calibrated, and therefore produces `LogLoss` metrics. These metrics don't makes sense for the non calibrated trainers.
We should therefore add an evaluation method named `EvaluateNonCalibrated` for the non calibrated trainers.
Contributor guide
Assessment
This issue has not been assessed yet.