ML Anthology

L1 Regularized Projection Pursuit for Additive Model Learning

Xiao Zhang, Lin Liang, Xiaoou Tang, Heung-Yeung Shum
CVPR 2008
doi:10.1109/CVPR.2008.4587356 /cvpr/2008/zhang2008cvpr-l/

Abstract

In this paper, we present a L$_{1}$ regularized projection pursuit algorithm for additive model learning. Two new algorithms are developed for regression and classification respectively: sparse projection pursuit regression and sparse Jensen-Shannon Boosting. The introduced L$_{1}$ regularized projection pursuit encourages sparse solutions, thus our new algorithms are robust to overfitting and present better generalization ability especially in settings with many irrelevant input features and noisy data. To make the optimization with L$_{1}$ regularization more efficient, we develop an "informative feature first" sequential optimization algorithm. Extensive experiments demonstrate the effectiveness of our proposed approach.

CVPR Semantic Scholar

Cite

Text

Zhang et al. "L1 Regularized Projection Pursuit for Additive Model Learning." IEEE/CVF Conference on Computer Vision and Pattern Recognition, 2008. doi:10.1109/CVPR.2008.4587356

Markdown

[Zhang et al. "L1 Regularized Projection Pursuit for Additive Model Learning." IEEE/CVF Conference on Computer Vision and Pattern Recognition, 2008.](https://mlanthology.org/cvpr/2008/zhang2008cvpr-l/) doi:10.1109/CVPR.2008.4587356

BibTeX

@inproceedings{zhang2008cvpr-l,
  title     = {{L1 Regularized Projection Pursuit for Additive Model Learning}},
  author    = {Zhang, Xiao and Liang, Lin and Tang, Xiaoou and Shum, Heung-Yeung},
  booktitle = {IEEE/CVF Conference on Computer Vision and Pattern Recognition},
  year      = {2008},
  doi       = {10.1109/CVPR.2008.4587356},
  url       = {https://mlanthology.org/cvpr/2008/zhang2008cvpr-l/}
}