A distributional approach for soft clustering comparison and evaluation

DOI10.1007/978-3-031-17801-6_1MaRDI QIDQ6160925zbMATH OpenOpenAlexFDO

Authors Andrea Campagner, Davide Ciucci, T. Denœux

Publication date 2 June 2023

Published in Belief Functions: Theory and Applications (Search for Journal in Brave)

Full work available at URL https://arxiv.org/abs/2206.09827

zbMATH Keywords

soft clusteringevidential clusteringexternal validation

Mathematics Subject Classification ID

Classification and discrimination; cluster analysis (statistical aspects) (62H30) Pattern recognition, speech recognition (68T10) Reasoning under uncertainty in the context of artificial intelligence (68T37)

Abstract: The development of external evaluation criteria for soft clustering (SC) has received limited attention: existing methods do not provide a general approach to extend comparison measures to SC, and are unable to account for the uncertainty represented in the results of SC algorithms. In this article, we propose a general method to address these limitations, grounding on a novel interpretation of SC as distributions over hard clusterings, which we call emph{distributional measures}. We provide an in-depth study of complexity- and metric-theoretic properties of the proposed approach, and we describe approximation techniques that can make the calculations tractable. Finally, we illustrate our approach through a simple but illustrative experiment.

Recommendations

• Comparing hard and overlapping clusterings
• Comparing clusterings by the variation of information.
• sARI: a soft agreement measure for class partitions incorporating assignment probabilities
• Criteria for comparing clusterings
• Soft clustering: why and how-to

Cites work

• scientific article; zbMATH DE number 3793445 (Why is no real title available?)
• Clustering and aggregation of relational data with applications to image database categorization
• Decision-making with belief functions: a review
• ECM: An evidential version of the fuzzy c-means algorithm
• External Indices for Rough Clustering
• Information theoretic measures for clusterings comparison: variants, properties, normalization and correction for chance
• Interval set clustering of web users with rough \(K\)-means
• The complexity of computing metric distances between partitions

Cited in

(3)

• Soft clustering: why and how-to
• A distributional framework for evaluation, comparison and uncertainty quantification in soft clustering
• sARI: a soft agreement measure for class partitions incorporating assignment probabilities

This page was built for publication: A distributional approach for soft clustering comparison and evaluation

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q6160925)