Estimating Continuous Distributions in Bayesian Classifiers
George H. John, Pat Langley · arXiv (Cornell University) · 2013
When modeling a probability distribution with a Bayesian network, we are faced with the problem of how to handle continuous variables. Most previous work has either solved the problem by discretizing, or assumed that the data are generated by a single Gaussian. In this paper we abandon the normality assumption and instead use statistical methods for nonparametric density estimation. For a naive Bayesian classifier, we present experimental results on a variety of natural and artificial domains, comparing two methods of density estimation: assuming normality and modeling each conditional distribution with a single Gaussian; and using nonparametric kernel density estimation. We observe large reductions in error on several natural and artificial data sets, which suggests that kernel estimation is a useful tool for learning Bayesian models.
Cited by 2,885 works. doi:10.48550/arxiv.1302.4964 Open-access PDF
Authors
References
- Maximum Likelihood from Incomplete Data Via the EM Algorithm
- Density Estimation for Statistics and Data Analysis
- Generalized Additive Models.
- Modern Applied Statistics with S-Plus.
- A Bayesian Method for the Induction of Probabilistic Networks from Data
- The CN2 Induction Algorithm
- Irrelevant Features and the Subset Selection Problem
- AutoClass: A Bayesian Classification System
- An analysis of Bayesian classifiers
- Supervised and Unsupervised Discretization of Continuous Features