Tackling Initial Centroid of K-Means with Distance Part (DP-KMeans)

TL;DR

This paper addresses the challenge of selecting initial centroids in k-means clustering, proposing a novel distance-based method to improve clustering accuracy especially as the number of clusters grows.

Contribution

It introduces a new distance part approach for initial centroid selection in k-means, enhancing clustering performance over traditional methods.

Findings

Improved clustering accuracy with the proposed method

More stable initial centroid selection in large cluster scenarios

Reduced sensitivity to initial centroid choice

Abstract

The initial centroid is a fairly challenging problem in the k-means method because it can affect the clustering results. In addition, choosing the starting centroid of the cluster is not always appropriate, especially, when the number of groups increases.