Clustering Stability: An Overview

Abstract

A popular method for selecting the number of clusters is based on stability arguments: one chooses the number of clusters such that the corresponding clustering results are "most stable". In recent years, a series of papers has analyzed the behavior of this method from a theoretical point of view. However, the results are very technical and difficult to interpret for non-experts. In this paper we give a high-level overview about the existing literature on clustering stability. In addition to presenting the results in a slightly informal but accessible way, we relate them to each other and discuss their different implications.

References in corpus (1)

• How the initialization affects the stability of the k-means algorithm

Cited by in corpus (11)

• Self-Supervised Representation Learning: Introduction, Advances and Challenges
• Role Discovery in Networks
• Consistency under sampling of exponential random graph models
• An information theoretic framework for classifying exoplanetary system architectures
• European banks' business models and their credit risk: A cluster analysis in a high-dimensional context
• ElegansNet: a brief scientific report and initial experiments
• A Latent Parameter Node-Centric Model for Spatial Networks
• Optimality-Based Clustering: An Inverse Optimization Approach
• Between Plateaus and Slopes: A Data-Driven Exploration of Spectral Diversity Across Type IIP/L Supernovae
• Clustering with Confidence: Finding Clusters with Statistical Guarantees
• Good Clusterings Have Large Volume