ML Anthology

TaylorBoost: First and Second-Order Boosting Algorithms with Explicit Margin Control

Mohammad J. Saberian, Hamed Masnadi-Shirazi, Nuno Vasconcelos
CVPR 2011 pp. 2929-2934
doi:10.1109/CVPR.2011.5995605 /cvpr/2011/saberian2011cvpr-taylorboost/

Abstract

A new family of boosting algorithms, denoted Taylor-Boost, is proposed. It supports any combination of loss function and first or second order optimization, and includes classical algorithms such as AdaBoost, Gradient-Boost, or LogitBoost as special cases. Its restriction to the set of canonical losses makes it possible to have boosting algorithms with explicit margin control. A new large family of losses with this property, based on the set of cumulative distributions of zero mean random variables, is then proposed. A novel loss function in this family, the Laplace loss, is finally derived. The combination of this loss and second order TaylorBoost produces a boosting algorithm with explicit margin control.

PDF CVPR Semantic Scholar

Cite

Text

Saberian et al. "TaylorBoost: First and Second-Order Boosting Algorithms with Explicit Margin Control." IEEE/CVF Conference on Computer Vision and Pattern Recognition, 2011. doi:10.1109/CVPR.2011.5995605

Markdown

[Saberian et al. "TaylorBoost: First and Second-Order Boosting Algorithms with Explicit Margin Control." IEEE/CVF Conference on Computer Vision and Pattern Recognition, 2011.](https://mlanthology.org/cvpr/2011/saberian2011cvpr-taylorboost/) doi:10.1109/CVPR.2011.5995605

BibTeX

@inproceedings{saberian2011cvpr-taylorboost,
  title     = {{TaylorBoost: First and Second-Order Boosting Algorithms with Explicit Margin Control}},
  author    = {Saberian, Mohammad J. and Masnadi-Shirazi, Hamed and Vasconcelos, Nuno},
  booktitle = {IEEE/CVF Conference on Computer Vision and Pattern Recognition},
  year      = {2011},
  pages     = {2929-2934},
  doi       = {10.1109/CVPR.2011.5995605},
  url       = {https://mlanthology.org/cvpr/2011/saberian2011cvpr-taylorboost/}
}