Data Clustering and Visualization with Recursive Max k-Cut Algorithm

TL;DR

This paper introduces a recursive max k-cut algorithm based on semidefinite programming that improves data clustering accuracy and efficiency, demonstrated through extensive experiments on datasets grouped into three clusters.

Contribution

It presents a novel recursive modification to the max k-cut algorithm utilizing semidefinite programming for better clustering performance.

Findings

Enhanced clustering accuracy over traditional methods

Improved computational efficiency in clustering tasks

Effective grouping of datasets into three clusters

Abstract

In this article, we continue our analysis for a novel recursive modification to the Max $k$-Cut algorithm using semidefinite programming as its basis, offering an improved performance in vectorized data clustering tasks. Using a dimension relaxation method, we use a recursion method to enhance density of clustering results. Our methods provide advantages in both computational efficiency and clustering accuracy for grouping datasets into three clusters, substantiated through comprehensive experiments.