File Exchange

image thumbnail

Dirichlet Process Gaussian Mixture Model

version 1.0 (6.29 KB) by

Dirichlet Process Gaussian Mixture Model aka Infinite GMM using Gibbs Sampling

34 Downloads

Updated

View License

This package solves the Dirichlet Process Gaussian Mixture Model (aka Infinite GMM) with Gibbs sampling. This is nonparametric Bayesian treatment for mixture model problems which automatically selects the proper number of the clusters.
I includes the Gaussian component distribution in the package. However, the code is flexible enough for Dirichlet process mixture model of any distribution. User can write your own class for the base distribution then let the underlying Gibbs sampling engine do the inference work.
Please try the demo script in the package.

This package is now a part of the PRML toolbox (http://www.mathworks.com/matlabcentral/fileexchange/55826-pattern-recognition-and-machine-learning-toolbox).

Comments and Ratings (6)

Jack Ma

Is there any detail description of the algorithm, just like you did for Variational Bayesian Inference for Gaussian Mixture Model?

Tiehang Duan

Yongsheng Li

yuebin wang

Rini

Is it applicable for 1D data as well?

Updates

1.0

update description

MATLAB Release
MATLAB 9.0 (R2016a)

Download apps, toolboxes, and other File Exchange content using Add-On Explorer in MATLAB.

» Watch video