Title :
Development and evaluation of kernel-based clustering validity indices
Author :
Fa, Rui ; Nandi, Asoke K. ; Abu-Jamous, Basel
Author_Institution :
Dept. of Electr. Eng. & Electron., Signal Process. & Commun. Res. Group, Univ. of Liverpool, Liverpool, UK
Abstract :
In this paper, there are two objectives: on one hand, we extend four conventional validity indices, namely the DI, the II, the CH, and the GI, to four kernel-based validity indices, correspondingly, the kDI, the kII, the kCH and the kGI; on the other hand, we conduct a Monte-Carlo simulation to evaluate and compare these validity indices. The numerical results show that some kernel validity indices work significant better than conventional ones and some of the validity indices work poorly or do not work at all in our study.
Keywords :
Monte Carlo methods; pattern clustering; unsupervised learning; Monte-Carlo simulation; kCH kernel-based validity index; kDI kernel-based validity index; kGI kernel-based validity index; kII kernel-based validity index; kernel-based clustering validity indices; unsupervised learning; Algorithm design and analysis; Clustering algorithms; Gene expression; Indexes; Kernel; Noise level; Standards;
Conference_Titel :
Signal Processing Conference (EUSIPCO), 2012 Proceedings of the 20th European
Conference_Location :
Bucharest
Print_ISBN :
978-1-4673-1068-0
