-
Notifications
You must be signed in to change notification settings - Fork 3.8k
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.
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
Piecewise linear trees #1315
Comments
Closed in favor of being in #2302. We decided to keep all feature requests in one place. Welcome to contribute this feature! Please re-open this issue (or post a comment if you are not a topic starter) if you are actively working on implementing this feature. |
I'd like to work on this issue. |
Thanks @kruda ! I'll re-open and assign to you. Let us know if you have any questions. |
@jameslamb I believe we should re-open a feature request issue only after any evidence of active work on it. Otherwise, we may fall down into dozens of open stale issues like it was before #2302, e.g. here #261 (comment). @kruda You can submit a PR any time you want marking it |
ok fair enough! |
Hi @kruda How is your process going? |
Implemented in #3299. |
I recently discovered this paper which was submitted earlier this year showing impressive results of GBDTs utilizing piecewise linear trees. It's claimed that PL trees converge faster and are more accurate than the constant trees xgboost & lightgbm uses, and that optimizations exist to make GBDT-PL competitive speed-wise.
The repo for the paper can be found here. Is there any chance of adding this feature?
The text was updated successfully, but these errors were encountered: