jmlr jmlr2008 jmlr2008-68 jmlr2008-68-reference knowledge-graph by maker-knowledge-mining
Source: pdf
Author: Eric Bax
Abstract: This paper develops bounds on out-of-sample error rates for support vector machines (SVMs). The bounds are based on the numbers of support vectors in the SVMs rather than on VC dimension. The bounds developed here improve on support vector counting bounds derived using Littlestone and Warmuth’s compression-based bounding technique. Keywords: compression, error bound, support vector machine, nearly uniform