An Uncertainty Framework for Classification

Abstract

We define a generalized likelihood function based on uncertainty measures and show that maximizing such a likelihood function for different measures induces different types of classifiers. In the probabilistic framework, we obtain classifiers that optimize the cross-entropy function. In the possibilistic framework, we obtain classifiers that maximize the interclass margin. Furthermore, we show that the support vector machine is a sub-class of these maximummargin classifiers.

Cite

Text

Teow and Loe. "An Uncertainty Framework for Classification." Conference on Uncertainty in Artificial Intelligence, 2000.

Markdown

[Teow and Loe. "An Uncertainty Framework for Classification." Conference on Uncertainty in Artificial Intelligence, 2000.](https://mlanthology.org/uai/2000/teow2000uai-uncertainty/)

BibTeX

@inproceedings{teow2000uai-uncertainty,
  title     = {{An Uncertainty Framework for Classification}},
  author    = {Teow, Loo-Nin and Loe, Kia-Fock},
  booktitle = {Conference on Uncertainty in Artificial Intelligence},
  year      = {2000},
  pages     = {574-579},
  url       = {https://mlanthology.org/uai/2000/teow2000uai-uncertainty/}
}