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 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.
Campagner, A., Ciucci, D., Denœux, T. (2022). A Distributional Approach for Soft Clustering Comparison and Evaluation. In 7th International Conference, BELIEF 2022, Paris, France, October 26–28, 2022, Proceedings (pp.3-12). Springer Science and Business Media Deutschland GmbH [10.1007/978-3-031-17801-6_1].
A Distributional Approach for Soft Clustering Comparison and Evaluation
Campagner, Andrea
Primo
;Ciucci, Davide;
2022
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 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.
I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.
