Bayesian Analysis

Default priors for density estimation with mixture models

J. E. Griffin

Full-text: Open access

Abstract

The infinite mixture of normals model has become a popular method for density estimation problems. This paper proposes an alternative hierarchical model that leads to hyperparameters that can be interpreted as the location, scale and smoothness of the density. The priors on other parts of the model have little effect on the density estimates and can be given default choices. Automatic Bayesian density estimation can be implemented by using uninformative priors for location and scale and default priors for the smoothness. The performance of these methods for density estimation are compared to previously proposed default priors for four data sets.

Article information

Source
Bayesian Anal., Volume 5, Number 1 (2010), 45-64.

Dates
First available in Project Euclid: 22 June 2012

Permanent link to this document
https://projecteuclid.org/euclid.ba/1340369792

Digital Object Identifier
doi:10.1214/10-BA502

Mathematical Reviews number (MathSciNet)
MR2596435

Zentralblatt MATH identifier
1330.62127

Keywords
Density Estimation Dirichlet process mixture models Mixtures of normals Normalized Generalized Gamma processes

Citation

Griffin, J. E. Default priors for density estimation with mixture models. Bayesian Anal. 5 (2010), no. 1, 45--64. doi:10.1214/10-BA502. https://projecteuclid.org/euclid.ba/1340369792


Export citation