Fitting Gaussian Mixtures by Automatic RJMCMC CPSC535c Course Project
نویسنده
چکیده
In this work, I will apply a recently developed automatic reversible jump MCMC algorithm to the problem of fitting a Gaussian mixture model to data. This algorithm, named AutoMix, is due to Hastie [5]. Reversible jump is approriate for this problem, because I will allow the number of mixture components to vary, and will seek to estimate the posterior distribution over both this quantity and the corresponding mixture parameters. Supposedly, Hastie’s academic supervisor, Green, was surprised that he did not apply AutoMix to the mixture of Gaussians problem. However, as my results will show, it is no surprise, as his algorithm is poorly suited to this task.
منابع مشابه
Robust Estimation in Gaussian Mixtures Using Multiresolution Kd-trees
For many applied problems in the context of clustering via mixture models, the estimates of the component means and covariance matrices can be affected by observations that are atypical of the components in the mixture model being fitted. In this paper, we consider for Gaussian mixtures a robust estimation procedure using multiresolution kd-trees. The method provides a fast EM-based approach to...
متن کاملSpeech Recognition Using Deep Learning Algorithms
Automatic speech recognition, translating of spoken words into text, is still a challenging task due to the high viability in speech signals. Deep learning, sometimes referred as representation learning or unsupervised feature learning, is a new area of machine learning. Deep learning is becoming a mainstream technology for speech recognition and has successfully replaced Gaussian mixtures for ...
متن کاملMCMC Methods for MLP-network and Gaussian Process and Stuff– A documentation for Matlab Toolbox MCMCstuff
MCMCstuff toolbox is a collection of Matlab functions for Bayesian inference with Markov chain Monte Carlo (MCMC) methods. This documentation introduces some of the features available in the toolbox. Introduction includes demonstrations of using Bayesian Multilayer Perceptron (MLP) network and Gaussian process in simple regression and classification problems with a hierarchical automatic releva...
متن کاملConvergence Assessment for Reversible Jump MCMC Simulations
In this paper we introduce the problem of assessing convergence of reversible jump MCMC algorithms on the basis of simulation output. We discuss the various direct approaches which could be employed, together with their associated drawbacks. Using the example of fitting a graphical Gaussian model via RJMCMC, we show how the simulation output for models which can be parameterised so that paramet...
متن کاملPhotometric method for the quantification of chlorophylls and their derivatives in complex mixtures: fitting with Gauss-peak spectra.
Accurate quantification of pigments in mixtures is essential in all cases in which separation of pigments by chromatography is impracticable for one reason or another. An example is the analysis of in vivo formation of heavy metal-substituted chlorophylls in heavy metal-stressed plants. We describe here a novel, accurate UV/VIS spectrophotometric method for the quantification of individual chlo...
متن کامل