Performance Analysis of Uncertain K-means Clustering Algorithm Using Different Distance Metrics

Autor:	Nitika Agarwal, Monal Jain, Swati Aggarwal
Rok vydání:	2018
Předmět:	Basis (linear algebra) Computer science Closeness k-means clustering Cluster (physics) Probability density function Data mining Cluster analysis computer.software_genre Measure (mathematics) computer
Zdroj:	Computational Intelligence: Theories, Applications and Future Directions-Volume I ISBN: 9789811311314
DOI:	10.1007/978-981-13-1132-1_19
Popis:	Real-world data generally deals with inconsistency. The uncertain k-means (UK-means) clustering algorithm, a modification of k-means, handles uncertain objects whose positions are represented by probability density functions (pdfs). Various techniques have been developed to enhance the performance of UK-means clustering algorithm but they are all centered on two major factors: choosing initial cluster centers and determining the number of clusters. This paper proposes that the measure of “closeness” is also a critical factor in deciding the quality of clusters. In this paper, the authors study the performance of UK-means clustering algorithm on four different distance functions using Haberman’s survival dataset. The analysis is performed on the basis of Davies–Bouldin index and purity values.
Databáze:	OpenAIRE
Externí odkaz:	https://explore.openaire.eu/search/publication?articleId=doi_________::4c973c2f3b23c2e9e3d61cc734e1dff3 https://doi.org/10.1007/978-981-13-1132-1_19 Zobrazit plný text záznamu