Clustering plays a fundamental role in Machine Learning. With clustering we refer to the problem of finding coherent groups in a dataset of elements. There are several algorithms to perform clustering that have been proposed in the literature, considering different costs for the optimization problems they consider. In this thesis we study the problem of clustering when the cost function is the silhouette coefficient, an index traditionally used for the internal validation of the results.

On the use of Silhouette for cost based clustering

Sansoni, Marco
2019/2020

Abstract

Clustering plays a fundamental role in Machine Learning. With clustering we refer to the problem of finding coherent groups in a dataset of elements. There are several algorithms to perform clustering that have been proposed in the literature, considering different costs for the optimization problems they consider. In this thesis we study the problem of clustering when the cost function is the silhouette coefficient, an index traditionally used for the internal validation of the results.
2019-10-15
clustering, silhouette, cost-based, heuristic, algorithm, k-means
File in questo prodotto:
File Dimensione Formato  
dissertation.pdf

accesso aperto

Dimensione 1.44 MB
Formato Adobe PDF
1.44 MB Adobe PDF Visualizza/Apri

The text of this website © Università degli studi di Padova. Full Text are published under a non-exclusive license. Metadata are under a CC0 License

Utilizza questo identificativo per citare o creare un link a questo documento: https://hdl.handle.net/20.500.12608/24616