Robust LogitBoost and Adaptive Base Class (ABC) LogitBoost

Ping Li
Proceedings of the 26th Conference on Uncertainty in Artificial Intelligence, PMLR R8:329-338, 2010.

Abstract

Logitboost is an influential boosting algorithm for classification. In this paper, we develop robust logitboost to provide an explicit formulation of tree-split criterion for building weak learners (regression trees) for logitboost. This formulation leads to a numerically stable implementation of logitboost. We then propose abc-logitboost for multi-class classification, by combining robust logitboost with the prior work of abc-boost. Previously, abc-boost was implemented as abc-mart using the mart algorithm. Our extensive experiments on multi-class classification compare four algorithms: mart, abc-mart, (robust) logitboost, and abc-logitboost, and demonstrate the superiority of abc-logitboost. Comparisons with other learning methods including SVM and deep learning are also available through prior publications.

Cite this Paper

BibTeX
@InProceedings{pmlr-vR8-li10c, title = {Robust LogitBoost and Adaptive Base Class ({ABC}) LogitBoost}, author = {Li, Ping}, booktitle = {Proceedings of the 26th Conference on Uncertainty in Artificial Intelligence}, pages = {329--338}, year = {2010}, editor = {Grünwald, Peter and Spirtes, Peter}, volume = {R8}, series = {Proceedings of Machine Learning Research}, month = {08--11 Jul}, publisher = {PMLR}, pdf = {https://raw.githubusercontent.com/mlresearch/r8/main/assets/li10c/li10c.pdf}, url = {https://proceedings.mlr.press/r8/li10c.html}, abstract = {Logitboost is an influential boosting algorithm for classification. In this paper, we develop robust logitboost to provide an explicit formulation of tree-split criterion for building weak learners (regression trees) for logitboost. This formulation leads to a numerically stable implementation of logitboost. We then propose abc-logitboost for multi-class classification, by combining robust logitboost with the prior work of abc-boost. Previously, abc-boost was implemented as abc-mart using the mart algorithm. Our extensive experiments on multi-class classification compare four algorithms: mart, abc-mart, (robust) logitboost, and abc-logitboost, and demonstrate the superiority of abc-logitboost. Comparisons with other learning methods including SVM and deep learning are also available through prior publications.}, note = {Reissued by PMLR on 04 October 2026.} }
Endnote
%0 Conference Paper %T Robust LogitBoost and Adaptive Base Class (ABC) LogitBoost %A Ping Li %B Proceedings of the 26th Conference on Uncertainty in Artificial Intelligence %C Proceedings of Machine Learning Research %D 2010 %E Peter Grünwald %E Peter Spirtes %F pmlr-vR8-li10c %I PMLR %P 329--338 %U https://proceedings.mlr.press/r8/li10c.html %V R8 %X Logitboost is an influential boosting algorithm for classification. In this paper, we develop robust logitboost to provide an explicit formulation of tree-split criterion for building weak learners (regression trees) for logitboost. This formulation leads to a numerically stable implementation of logitboost. We then propose abc-logitboost for multi-class classification, by combining robust logitboost with the prior work of abc-boost. Previously, abc-boost was implemented as abc-mart using the mart algorithm. Our extensive experiments on multi-class classification compare four algorithms: mart, abc-mart, (robust) logitboost, and abc-logitboost, and demonstrate the superiority of abc-logitboost. Comparisons with other learning methods including SVM and deep learning are also available through prior publications. %Z Reissued by PMLR on 04 October 2026.
APA
Li, P.. (2010). Robust LogitBoost and Adaptive Base Class (ABC) LogitBoost. Proceedings of the 26th Conference on Uncertainty in Artificial Intelligence, in Proceedings of Machine Learning Research R8:329-338 Available from https://proceedings.mlr.press/r8/li10c.html. Reissued by PMLR on 04 October 2026.

Related Material