This paper presents a novel fuzzy clustering technique designed specifically for count data, referred to as the Fuzzy C-medoids algorithm based on Total Variation Distance. We evaluate its performance against a benchmark relying on Shannon divergence, commonly employed in scenarios involving discrete probability distributions, through simulation analysis. A comprehensive evaluation of the proposed approach’s effectiveness is carried out, revealing promising results. The study’s findings emphasize the potential of the proposed fuzzy method, particularly in scenarios where discrete probability distributions are involved.

Copula-Based Fuzzy Clustering of Count Data with Total Variation Distance / D'Urso, Pierpaolo; De Giovanni, Livia; Federico, Lorenzo; Vitale, Vincenzina. - (2024), pp. 126-133. [10.1007/978-3-031-65993-5_15]

Copula-Based Fuzzy Clustering of Count Data with Total Variation Distance

Pierpaolo D'Urso;Livia De Giovanni;Lorenzo Federico
;
Vincenzina Vitale
2024

Abstract

This paper presents a novel fuzzy clustering technique designed specifically for count data, referred to as the Fuzzy C-medoids algorithm based on Total Variation Distance. We evaluate its performance against a benchmark relying on Shannon divergence, commonly employed in scenarios involving discrete probability distributions, through simulation analysis. A comprehensive evaluation of the proposed approach’s effectiveness is carried out, revealing promising results. The study’s findings emphasize the potential of the proposed fuzzy method, particularly in scenarios where discrete probability distributions are involved.
2024
978-3-031-65992-8
Total Variation Distance, Fuzzy C-medoids, Count data
Copula-Based Fuzzy Clustering of Count Data with Total Variation Distance / D'Urso, Pierpaolo; De Giovanni, Livia; Federico, Lorenzo; Vitale, Vincenzina. - (2024), pp. 126-133. [10.1007/978-3-031-65993-5_15]
File in questo prodotto:
File Dimensione Formato  
Copula_based_fuzzy_clustering_of_count_data_with_Total_Variation_distance.pdf

Solo gestori archivio

Tipologia: Documento in Post-print
Licenza: Tutti i diritti riservati
Dimensione 242.1 kB
Formato Adobe PDF
242.1 kB Adobe PDF   Visualizza/Apri
Pubblicazioni consigliate

I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.

Utilizza questo identificativo per citare o creare un link a questo documento: https://hdl.handle.net/11385/240578
Citazioni
  • Scopus ND
  • ???jsp.display-item.citation.isi??? ND
  • OpenAlex ND
social impact