Clustering, factor discovery and optimal transport

DOI10.1093/IMAIAI/IAAA040MaRDI QIDQ5033277zbMATH OpenOpenAlexFDO

Authors Hongkang Yang, Esteban G. Tabak

Publication date 22 February 2022

Published in Information and Inference: A Journal of the IMA (Search for Journal in Brave)

Full work available at URL https://arxiv.org/abs/1902.10288

zbMATH Keywords

clusteringexplanation of variabilityfactor discoveryoptimal transportprincipal curveWasserstein barycenter

Mathematics Subject Classification ID

Classification and discrimination; cluster analysis (statistical aspects) (62H30)

Abstract: The clustering problem, and more generally, latent factor discovery --or latent space inference-- is formulated in terms of the Wasserstein barycenter problem from optimal transport. The objective proposed is the maximization of the variability attributable to class, further characterized as the minimization of the variance of the Wasserstein barycenter. Existing theory, which constrains the transport maps to rigid translations, is extended to affine transformations. The resulting non-parametric clustering algorithms include k-means as a special case and exhibit more robust performance. A continuous version of these algorithms discovers continuous latent variables and generalizes principal curves. The strength of these algorithms is demonstrated by tests on both artificial and real-world data sets.

Recommendations

• Robust clustering tools based on optimal transportation
• On the computation of Wasserstein barycenters
• Hierarchical clustering with optimal transport
• Statistical data analysis in the Wasserstein space
• Hybrid Wasserstein distance and fast distribution clustering

Cited in

(8)

• Hierarchical clustering with optimal transport
• Hybrid Wasserstein distance and fast distribution clustering
• Robust clustering tools based on optimal transportation
• scientific article; zbMATH DE number 4036950 (Why is no real title available?)
• Explanation of variability and removal of confounding factors from data through optimal transport
• Wasserstein dictionary learning: optimal transport-based unsupervised nonlinear dictionary learning
• Optimal transport, mean partition, and uncertainty assessment in cluster analysis
• New algorithmic directions in optimal transport and applications for product spaces

This page was built for publication: Clustering, factor discovery and optimal transport

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q5033277)