Skip to content

OneVersusAll calibrator for naturally calibrated models #2805

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Jump to bottom
Open
singlis opened this issue Feb 28, 2019 · 0 comments
Open

OneVersusAll calibrator for naturally calibrated models #2805

singlis opened this issue Feb 28, 2019 · 0 comments
Labels
P3 Doc bugs, questions, minor issues, etc. question Further information is requested

Comments

@singlis
Copy link
Member

singlis commented Feb 28, 2019

This question came up during a code review of some changes to the One Versus All (OVA) Trainer. The OVA Trainer takes in a calibrator. If one is not specified, the PlattCalibrator is then used. After looking at the code, it looks like the calibrator is attached to each model instance that the OVA Trainer evaluates.

The question is what happens when you are using a naturally calibrated mode like logistic regression? Can this produce unexpected results?
@singlis singlis added the question Further information is requested label Feb 28, 2019
@singlis singlis mentioned this issue Feb 28, 2019
Merged
@harishsk harishsk added the P3 Doc bugs, questions, minor issues, etc. label Jan 12, 2020
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Assignees
No one assigned
Labels
P3 Doc bugs, questions, minor issues, etc. question Further information is requested
Projects
None yet
Milestone
No milestone
Development

No branches or pull requests
2 participants
@harishsk @singlis