ML Anthology

MKPM: A Multiclass Extension to the Kernel Projection Machine

Sylvain Takerkart, Liva Ralaivola
CVPR 2011 pp. 2785-2791
doi:10.1109/CVPR.2011.5995657 /cvpr/2011/takerkart2011cvpr-mkpm/

Abstract

We introduce Multiclass Kernel Projection Machines (MKPM), a new formalism that extends the Kernel Projection Machine framework to the multiclass case. Our formulation is based on the use of output codes and it implements a co-regularization scheme by simultaneously constraining the projection dimensions associated with the individual predictors that constitute the global classifier. In order to solve the optimization problem posed by our formulation, we propose an efficient dynamic programming approach. Numerical simulations conducted on a few pattern recognition problems illustrate the soundness of our approach.

PDF CVPR Semantic Scholar

Cite

Text

Takerkart and Ralaivola. "MKPM: A Multiclass Extension to the Kernel Projection Machine." IEEE/CVF Conference on Computer Vision and Pattern Recognition, 2011. doi:10.1109/CVPR.2011.5995657

Markdown

[Takerkart and Ralaivola. "MKPM: A Multiclass Extension to the Kernel Projection Machine." IEEE/CVF Conference on Computer Vision and Pattern Recognition, 2011.](https://mlanthology.org/cvpr/2011/takerkart2011cvpr-mkpm/) doi:10.1109/CVPR.2011.5995657

BibTeX

@inproceedings{takerkart2011cvpr-mkpm,
  title     = {{MKPM: A Multiclass Extension to the Kernel Projection Machine}},
  author    = {Takerkart, Sylvain and Ralaivola, Liva},
  booktitle = {IEEE/CVF Conference on Computer Vision and Pattern Recognition},
  year      = {2011},
  pages     = {2785-2791},
  doi       = {10.1109/CVPR.2011.5995657},
  url       = {https://mlanthology.org/cvpr/2011/takerkart2011cvpr-mkpm/}
}