A Sparse PCA Approach to Clustering

TL;DR

This paper introduces a clustering method based on sparse PCA for Gaussian mixture models, extending to cases with non-diagonal covariance matrices, and compares it with existing methods.

Contribution

It proposes a new sparse PCA-based clustering approach and analyzes its performance, including in dependent covariance scenarios, providing a comparative study.

Findings

Effective clustering with sparse PCA in Gaussian mixtures

Performance comparison with IF-PCA method

Extension to dependent covariance matrices

Abstract

We discuss a clustering method for Gaussian mixture model based on the sparse principal component analysis (SPCA) method and compare it with the IF-PCA method. We also discuss the dependent case where the covariance matrix $\Sigma$ is not necessarily diagonal.