Similarity-Driven Cluster Merging Method for Unsupervised Fuzzy Clustering

TL;DR

This paper introduces a similarity-driven cluster merging approach for unsupervised fuzzy clustering that automatically determines the number of clusters and improves clustering validation using a modified objective function.

Contribution

It presents a novel cluster merging criterion based on fuzzy similarity and an adaptive threshold, along with a modified objective function incorporating principal components.

Findings

Effective in resolving cluster validation issues

Automatically determines the optimal number of clusters

Demonstrates improved clustering performance in experiments

Abstract

In this paper, a similarity-driven cluster merging method is proposed for unsuper-vised fuzzy clustering. The cluster merging method is used to resolve the problem of cluster validation. Starting with an overspecified number of clusters in the data, pairs of similar clusters are merged based on the proposed similarity-driven cluster merging criterion. The similarity between clusters is calculated by a fuzzy cluster similarity matrix, while an adaptive threshold is used for merging. In addition, a modified generalized ob- jective function is used for prototype-based fuzzy clustering. The function includes the p-norm distance measure as well as principal components of the clusters. The number of the principal components is determined automatically from the data being clustered. The properties of this unsupervised fuzzy clustering algorithm are illustrated by several experiments.