Abstract:Mixture models with Gamma and or inverse-Gamma distributed mixture components are useful for medical image tissue segmentation or as post-hoc models for regression coefficients obtained from linear regression within a Generalised Linear Modeling framework (GLM), used in this case to separate stochastic (Gaussian) noise from some kind of positive or negative "activation" (modeled as Gamma or inverse-Gamma distributed). To date, the most common choice in this context it is Gaussian/Gamma mixture models learned through a maximum likelihood (ML) approach; we recently extended such algorithm for mixture models with inverse-Gamma components. Here, we introduce a fully analytical Variational Bayes (VB) learning framework for both Gamma and/or inverse-Gamma components. We use synthetic and resting state fMRI data to compare the performance of the ML and VB algorithms in terms of area under the curve and computational cost. We observed that the ML Gaussian/Gamma model is very expensive specially when considering high resolution images; furthermore, these solutions are highly variable and they occasionally can overestimate the activations severely. The Bayesian Gauss-Gamma is in general the fastest algorithm but provides too dense solutions. The maximum likelihood Gaussian/inverse-Gamma is also very fast but provides in general very sparse solutions. The variational Gaussian/inverse-Gamma mixture model is the most robust and its cost is acceptable even for high resolution images. Further, the presented methodology represents an essential building block that can be directly used in more complex inference tasks, specially designed to analyse MRI-fMRI data; such models include for example analytical variational mixture models with adaptive spatial regularization or better source models for new spatial blind source separation approaches.

Variational Learning for Finite Beta-Liouville Mixture Models and Its Application

Variational Learning for Finite Inverted Dirichlet Mixture Models and Applications

Bayesian Estimation of Beta Mixture Models with Variational Inference

Bayesian estimation and prediction for certain mixtures

Extended Variational Inference for Gamma Mixture Model in Positive Vectors Modeling

Extended Variational Inference for Dirichlet Process Mixture of Beta‐Liouville Distributions for Proportional Data Modeling

Parameter Estimation of Gaussian Mixture Model Based on Variational Bayesian Learning

MAP approximation to the variational Bayes Gaussian mixture model and application

Bayesian Estimation of Inverted Beta Mixture Models with Extended Stochastic Variational Inference for Positive Vector Classification.

Variational Bayesian Learning for Parameter Estimation of Mixture of Gaussians

Variational Bayesian Learning for Dirichlet Process Mixture of Inverted Dirichlet Distributions in Non-Gaussian Image Feature Modeling

Variational Mixture Models with Gamma or inverse-Gamma components

Efficient Variational Bayesian Model Updating by Bayesian Active Learning

Bayesian Estimation of the Von-Mises Fisher Mixture Model with Variational Inference

Efficient Mixture Learning in Black-Box Variational Inference

Memorized Variational Continual Learning for Dirichlet Process Mixtures

Predictive Distribution of the Dirichlet Mixture Model by Local Variational Inference

The Infinite Student'S T-Mixture For Robust Modeling

Positive Data Modeling Using a Mixture of Mixtures of Inverted Beta Distributions.

Expectation propagation learning of finite and infinite Gamma mixture models and its applications

Novel mixture allocation models for topic learning