Discriminative Model Selection for Density Models
Abstract
Density models are a popular tool for building classifiers. When using density models to build a classifier, one typically learns a separate density model for each class of interest. These density models are then combined to make a classifier through the use of Bayes’ rule utilizing the prior distribution over the classes. In this paper, we provide a discriminative method for choosing among alternative density models for each class to improve classification accuracy.
Cite
Text
Thiesson and Meek. "Discriminative Model Selection for Density Models." Proceedings of the Ninth International Workshop on Artificial Intelligence and Statistics, 2003.Markdown
[Thiesson and Meek. "Discriminative Model Selection for Density Models." Proceedings of the Ninth International Workshop on Artificial Intelligence and Statistics, 2003.](https://mlanthology.org/aistats/2003/thiesson2003aistats-discriminative/)BibTeX
@inproceedings{thiesson2003aistats-discriminative,
title = {{Discriminative Model Selection for Density Models}},
author = {Thiesson, Bo and Meek, Christopher},
booktitle = {Proceedings of the Ninth International Workshop on Artificial Intelligence and Statistics},
year = {2003},
pages = {270-275},
volume = {R4},
url = {https://mlanthology.org/aistats/2003/thiesson2003aistats-discriminative/}
}