ML Anthology

Unifying the Error-Correcting and Output-Code AdaBoost Within the Margin Framework

Yijun Sun, Sinisa Todorovic, Jian Li, Dapeng Wu
ICML 2005 pp. 872-879
doi:10.1145/1102351.1102461 /icml/2005/sun2005icml-unifying/

Abstract

In this paper, we present a new interpretation of AdaBoost. ECC and AdaBoost. OC. We show that AdaBoost. ECC performs stage-wise functional gradient descent on a cost function, defined in the domain of margin values, and that AdaBoost. OC is a shrinkage version of AdaBoost. ECC. These findings strictly explain some properties of the two algorithms. The gradient-minimization formulation of AdaBoost. ECC allows us to derive a new algorithm, referred to as AdaBoost. SECC, by explicitly exploiting shrinkage as regularization in AdaBoost. ECC. Experiments on diverse databases confirm our theoretical findings. Empirical results show that AdaBoost. SECC performs significantly better than AdaBoost. ECC and AdaBoost. OC.

PDF ICML Semantic Scholar

Cite

Text

Sun et al. "Unifying the Error-Correcting and Output-Code AdaBoost Within the Margin Framework." International Conference on Machine Learning, 2005. doi:10.1145/1102351.1102461

Markdown

[Sun et al. "Unifying the Error-Correcting and Output-Code AdaBoost Within the Margin Framework." International Conference on Machine Learning, 2005.](https://mlanthology.org/icml/2005/sun2005icml-unifying/) doi:10.1145/1102351.1102461

BibTeX

@inproceedings{sun2005icml-unifying,
  title     = {{Unifying the Error-Correcting and Output-Code AdaBoost Within the Margin Framework}},
  author    = {Sun, Yijun and Todorovic, Sinisa and Li, Jian and Wu, Dapeng},
  booktitle = {International Conference on Machine Learning},
  year      = {2005},
  pages     = {872-879},
  doi       = {10.1145/1102351.1102461},
  url       = {https://mlanthology.org/icml/2005/sun2005icml-unifying/}
}