Mixtures of Common Skew-t Factor Analyzers

TL;DR

This paper introduces a novel mixture of common skew-t factor analyzers model for robust, high-dimensional clustering, effectively handling skewed data and large numbers of components with improved performance.

Contribution

It is the first to incorporate skewed common factors into a mixture of factor analyzers, enhancing clustering robustness and scalability.

Findings

Excellent clustering performance on real and simulated data

Effective handling of skewed high-dimensional data

Robustness to a large number of mixture components

Abstract

A mixture of common skew-t factor analyzers model is introduced for model-based clustering of high-dimensional data. By assuming common component factor loadings, this model allows clustering to be performed in the presence of a large number of mixture components or when the number of dimensions is too large to be well-modelled by the mixtures of factor analyzers model or a variant thereof. Furthermore, assuming that the component densities follow a skew-t distribution allows robust clustering of skewed data. The alternating expectation-conditional maximization algorithm is employed for parameter estimation. We demonstrate excellent clustering performance when our model is applied to real and simulated data.This paper marks the first time that skewed common factors have been used.