Regularization and Optimization in Model-Based Clustering

TL;DR

This paper introduces improved optimization and regularization techniques for Gaussian Mixture Models, significantly enhancing cluster recovery and revealing complex data structures beyond traditional k-means methods.

Contribution

It develops new algorithms and regularization strategies for GMMs, enabling better clustering performance and overfitting prevention, with open-source Julia packages provided.

Findings

Combining optimization and regularization yields superior clustering results.

Enhanced algorithms outperform existing methods in recovering complex clusters.

Open-source tools facilitate broader application of advanced GMM techniques.

Abstract

Due to their conceptual simplicity, k-means algorithm variants have been extensively used for unsupervised cluster analysis. However, one main shortcoming of these algorithms is that they essentially fit a mixture of identical spherical Gaussians to data that vastly deviates from such a distribution. In comparison, general Gaussian Mixture Models (GMMs) can fit richer structures but require estimating a quadratic number of parameters per cluster to represent the covariance matrices. This poses two main issues: (i) the underlying optimization problems are challenging due to their larger number of local minima, and (ii) their solutions can overfit the data. In this work, we design search strategies that circumvent both issues. We develop more effective optimization algorithms for general GMMs, and we combine these algorithms with regularization strategies that avoid overfitting. Through extensive computational analyses, we observe that optimization or regularization in isolation does not substantially improve cluster recovery. However, combining these techniques permits a completely new level of performance previously unachieved by k-means algorithm variants, unraveling vastly different cluster structures. These results shed new light on the current status quo between GMM and k-means methods and suggest the more frequent use of general GMMs for data exploration. To facilitate such applications, we provide open-source code as well as Julia packages (UnsupervisedClustering.jl and RegularizedCovarianceMatrices.jl) implementing the proposed techniques.