Abstract
In this paper, we propose a modified sequential minimal optimization (SMO) algorithm for the minimal enclosing sphere estimation (MESE) problem. Being one of the key issues of the VC dimension estimation, the MESE problem has a formulation similar to that of support vector machines (SVMs) training. This allows adoption of the ideas of Platt's SMO algorithm. After careful analysis of the MESE problem, key issues are addressed. Experimental results show the feasibility and effectiveness of the proposed algorithm when applied to SVMs model selection. © Springer-Verlag 2004.
Cite
CITATION STYLE
Li, H., Wang, S., & Qi, F. (2004). Minimal enclosing sphere estimation and its application to SVMs model selection. Lecture Notes in Computer Science (Including Subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), 3173, 487–493. https://doi.org/10.1007/978-3-540-28647-9_81
Register to see more suggestions
Mendeley helps you to discover research relevant for your work.
