Stop using the elbow criterion for k-means and how to choose the number of clusters instead

Abstract

A major challenge when using k-means clustering often is how to choose the parameter k, the number of clusters. In this letter, we want to point out that it is very easy to draw poor conclusions from a common heuristic, the "elbow method". Better alternatives have been known in literature for a long time, and we want to draw attention to some of these easy to use options, that often perform better. This letter is a call to stop using the elbow method altogether, because it severely lacks theoretic support, and we want to encourage educators to discuss the problems of the method -- if introducing it in class at all -- and teach alternatives instead, while researchers and reviewers should reject conclusions drawn from the elbow method.

References in corpus (2)

• Clustering by Direct Optimization of the Medoid Silhouette
• Accelerating Spherical k-Means

Cited by in corpus (8)

• Medoid Silhouette clustering with automatic cluster number selection
• CARL-G: Clustering-Accelerated Representation Learning on Graphs
• Keeping Score: A Quantitative Analysis of How the CHI Community Appreciates Its Milestones
• Discovering Operational Patterns Using Image-Based Convolutional Clustering and Composite Evaluation: A Case Study in Foundry Melting Processes
• Multivariate Time Series Clustering for Environmental State Characterization of Ground-Based Gravitational-Wave Detectors
• Accelerating k-Means Clustering with Cover Trees
• Similar Fermi-GBM sGRBs to GW/sGRB 170817A in MeV-GeV energies
• Resilience metrics to guide back-up investments in the power system during extreme weather