Naive Bayesian Classifiers with Extreme Probability Features

Linda C. van der Gaag, Andrea Capotorti
Proceedings of the Ninth International Conference on Probabilistic Graphical Models, PMLR 72:499-510, 2018.

Abstract

Despite their popularity, naive Bayesian classifiers are not well suited for real-world applications involving extreme probability features. As will be demonstrated in this paper, methods used to forestall the inclusion of zero probability parameters in naive classifiers have quite counterintuitive effects. An elegant, principled solution for handling extreme probability events is available however, from coherent conditional probability theory. We will show how this theory can be integrated in standard naive Bayesian classifiers, and then present a computational framework that retains the classifiers’ efficiency in the presence of a limited number of extreme probability features.

Cite this Paper

BibTeX
@InProceedings{pmlr-v72-van-der-gaag18b, title = {Naive Bayesian Classifiers with Extreme Probability Features}, author = {{van der Gaag}, Linda C. and Capotorti, Andrea}, booktitle = {Proceedings of the Ninth International Conference on Probabilistic Graphical Models}, pages = {499--510}, year = {2018}, editor = {Kratochvíl, Václav and Studený, Milan}, volume = {72}, series = {Proceedings of Machine Learning Research}, month = {11--14 Sep}, publisher = {PMLR}, pdf = {http://proceedings.mlr.press/v72/van-der-gaag18b/van-der-gaag18b.pdf}, url = {https://proceedings.mlr.press/v72/van-der-gaag18b.html}, abstract = {Despite their popularity, naive Bayesian classifiers are not well suited for real-world applications involving extreme probability features. As will be demonstrated in this paper, methods used to forestall the inclusion of zero probability parameters in naive classifiers have quite counterintuitive effects. An elegant, principled solution for handling extreme probability events is available however, from coherent conditional probability theory. We will show how this theory can be integrated in standard naive Bayesian classifiers, and then present a computational framework that retains the classifiers’ efficiency in the presence of a limited number of extreme probability features.} }
Endnote
%0 Conference Paper %T Naive Bayesian Classifiers with Extreme Probability Features %A Linda C. van der Gaag %A Andrea Capotorti %B Proceedings of the Ninth International Conference on Probabilistic Graphical Models %C Proceedings of Machine Learning Research %D 2018 %E Václav Kratochvíl %E Milan Studený %F pmlr-v72-van-der-gaag18b %I PMLR %P 499--510 %U https://proceedings.mlr.press/v72/van-der-gaag18b.html %V 72 %X Despite their popularity, naive Bayesian classifiers are not well suited for real-world applications involving extreme probability features. As will be demonstrated in this paper, methods used to forestall the inclusion of zero probability parameters in naive classifiers have quite counterintuitive effects. An elegant, principled solution for handling extreme probability events is available however, from coherent conditional probability theory. We will show how this theory can be integrated in standard naive Bayesian classifiers, and then present a computational framework that retains the classifiers’ efficiency in the presence of a limited number of extreme probability features.
APA
van der Gaag, L.C. & Capotorti, A.. (2018). Naive Bayesian Classifiers with Extreme Probability Features. Proceedings of the Ninth International Conference on Probabilistic Graphical Models, in Proceedings of Machine Learning Research 72:499-510 Available from https://proceedings.mlr.press/v72/van-der-gaag18b.html.

Related Material