Variational Bayesian Inference For A Scale Mixture Of Normal Distributions Handling Missing Data

TL;DR

This paper introduces a Variational Bayesian approach for a scale mixture of Normal distributions to improve classification and clustering robustness in datasets with outliers and missing data.

Contribution

It develops a novel mixture model with latent variables and Bayesian inference to better handle outliers and missing data in classification and clustering tasks.

Findings

Effective handling of outliers and missing data demonstrated

Bayesian inference improves model robustness

Enhanced classification and clustering performance

Abstract

In this paper, a scale mixture of Normal distributions model is developed for classification and clustering of data having outliers and missing values. The classification method, based on a mixture model, focuses on the introduction of latent variables that gives us the possibility to handle sensitivity of model to outliers and to allow a less restrictive modelling of missing data. Inference is processed through a Variational Bayesian Approximation and a Bayesian treatment is adopted for model learning, supervised classification and clustering.