Paper Info
Title
On Bayesian analysis of a finite generalized Dirichlet mixture via a Metropolis-within-Gibbs sampling
Abstract
In this paper, we present a fully Bayesian approach for generalized Dirichlet mixtures estimation and selection. The estimation of the parameters is based on the Monte Carlo simulation technique of Gibbs sampling mixed with a Metropolis-Hastings step. Also, we obtain a posterior distribution which is conjugate to a generalized Dirichlet likelihood. For the selection of the number of clusters, we used the integrated likelihood. The performance of our Bayesian algorithm is tested and compared with the maximum likelihood approach by the classification of several synthetic and real data sets. The generalized Dirichlet mixture is also applied to the problems of IR eye modeling and introduced as a probabilistic kernel for Support Vector Machines.
Year
DOI
Venue
2009
10.1007/s10044-008-0111-4
Pattern Anal. Appl.
Keywords
Field
DocType
generalized dirichlet mixtures estimation,metropolis-hastings step,bayesian algorithm,bayesian analysis,generalized dirichlet mixtureem � bayesian analysisgibbs samplingmetropolis-hastings � svmimages classification,ir eye modeling,metropolis-within-gibbs sampling,integrated likelihood,monte carlo simulation technique,bayesian approach,generalized dirichlet mixture,maximum likelihood approach,finite generalized dirichlet mixture,generalized dirichlet likelihood,gibbs sampling,support vector machine,metropolis hastings,monte carlo simulation,maximum likelihood,posterior distribution
Dirichlet-multinomial distribution,Latent Dirichlet allocation,Categorical distribution,Pattern recognition,Marginal likelihood,Generalized Dirichlet distribution,Artificial intelligence,Estimation theory,Dirichlet distribution,Gibbs sampling,Mathematics
Journal
Volume
Issue
ISSN
12
2
1433-755X
Citations 
PageRank 
References 
28
0.74
35
Authors
3
Name
Order
Citations
PageRank
Nizar Bouguila11539146.09
Djemel Ziou2139599.40
Riad I. Hammoud31189.46