THE CONNECTION BETWEEN REGULARIZATION OPERATORS AND SUPPORT VECTOR KERNELS

Citation
Aj. Smola et al., THE CONNECTION BETWEEN REGULARIZATION OPERATORS AND SUPPORT VECTOR KERNELS, Neural networks, 11(4), 1998, pp. 637-649
Citations number
39
Categorie Soggetti
Computer Science Artificial Intelligence","Computer Science Artificial Intelligence
Journal title
ISSN journal
08936080
Volume
11
Issue
4
Year of publication
1998
Pages
637 - 649
Database
ISI
SICI code
0893-6080(1998)11:4<637:TCBROA>2.0.ZU;2-C
Abstract
In this paper a correspondence is derived between regularization opera tors used in regularization networks and support vector kernels. We pr ove that the Green's Functions associated with regularization operator s are suitable support vector kernels with equivalent regularization p roperties. Moreover, the paper provides an analysis of currently used support vector kernels in the view of regularization theory and corres ponding operators associated with the classes of both polynomial kerne ls and translation invariant kernels. The latter are also analyzed on periodical domains. As a by-product we show that a large number of rad ial basis functions, namely conditionally positive definite functions, may be used as support vector kernels. (C) 1998 Elsevier Science Ltd. All rights reserved.