Generalized density clustering

TL;DR

This paper explores generalized density clustering, enabling accurate identification of clusters on lower-dimensional manifolds in high-dimensional data, with methods for bandwidth selection and stability analysis.

Contribution

It introduces a framework for density clustering on lower-dimensional structures and proposes practical algorithms for bandwidth choice and cluster approximation.

Findings

Accurate clustering on lower-dimensional manifolds in high dimensions.

Effective data-based bandwidth selection methods.

A simple graph-based algorithm approximates high-density clusters.

Abstract

We study generalized density-based clustering in which sharply defined clusters such as clusters on lower-dimensional manifolds are allowed. We show that accurate clustering is possible even in high dimensions. We propose two data-based methods for choosing the bandwidth and we study the stability properties of density clusters. We show that a simple graph-based algorithm successfully approximates the high density clusters.