dotnet / dotnet/machinelearning
Add Non-negative support for Averaged Perceptron
Open
area-Trainers
enhancement
Priority:2
- Dominant language
- C#
- Stars
- 9.4k
- Forks
- 2k
- Avg merge
- 2d 20h
- Merged PRs (30d)
- 11
Description
This is a request to add non-negative weight support to AveragedPerceptron, similar to Logistic Regression through the EnforceNonNegativity ("nn=+") argument.
The reason of asking for this is because Averaged Perceptron can at times perform better in terms of timing and accuracy -- therefore it would useful to have parity with this feature.
@justinormont
Contributor guide
Assessment
This issue has not been assessed yet.