Dissimilarity Clustering by Hierarchical Multi-Level Refinement

TL;DR

This paper presents a hierarchical multi-level refinement approach for dissimilarity clustering, improving quantization error optimization in k-means clustering for dissimilarity data.

Contribution

It introduces a novel hierarchical and heuristic refinement method that enhances clustering quality and computational efficiency for dissimilarity-based data.

Findings

Achieves lower quantization errors than existing methods

Computationally efficient for large datasets

Effective in optimizing dissimilarity clustering

Abstract

We introduce in this paper a new way of optimizing the natural extension of the quantization error using in k-means clustering to dissimilarity data. The proposed method is based on hierarchical clustering analysis combined with multi-level heuristic refinement. The method is computationally efficient and achieves better quantization errors than the