Support Vector Machines as Probabilistic Models
Abstract
We show how the SVM can be viewed as a maximum likelihood estimate of a class of probabilistic models. This model class can be viewed as a reparametrization of the SVM in a similar vein to the $\nu$-SVM reparametrizing the original SVM. It is not discriminative, but has a non-uniform marginal. We illustrate the benefits of this new view by re-deriving and re-investigating two established SVM-related algorithms.
Cite
Text
Franc et al. "Support Vector Machines as Probabilistic Models." International Conference on Machine Learning, 2011.Markdown
[Franc et al. "Support Vector Machines as Probabilistic Models." International Conference on Machine Learning, 2011.](https://mlanthology.org/icml/2011/franc2011icml-support/)BibTeX
@inproceedings{franc2011icml-support,
title = {{Support Vector Machines as Probabilistic Models}},
author = {Franc, Vojtech and Zien, Alexander and Schölkopf, Bernhard},
booktitle = {International Conference on Machine Learning},
year = {2011},
pages = {665-672},
url = {https://mlanthology.org/icml/2011/franc2011icml-support/}
}