Risk Bounds for Regularized Least-squares Algorithm with Operator-valued kernels

dc.contributor.author	Vito, Ernesto De
dc.contributor.author	Caponnetto, Andrea
dc.date.accessioned	2005-12-22T02:28:54Z
dc.date.available	2005-12-22T02:28:54Z
dc.date.issued	2005-05-16
dc.identifier.other	MIT-CSAIL-TR-2005-031
dc.identifier.other	AIM-2005-015
dc.identifier.other	CBCL-249
dc.identifier.uri	http://hdl.handle.net/1721.1/30543
dc.description.abstract	We show that recent results in [3] on risk bounds for regularized least-squares on reproducing kernel Hilbert spaces can be straightforwardly extended to the vector-valued regression setting. We first briefly introduce central concepts on operator-valued kernels. Then we show how risk bounds can be expressed in terms of a generalization of effective dimension.
dc.format.extent	17 p.
dc.format.extent	12090406 bytes
dc.format.extent	642646 bytes
dc.format.mimetype	application/postscript
dc.format.mimetype	application/pdf
dc.language.iso	en_US
dc.relation.ispartofseries	Massachusetts Institute of Technology Computer Science and Artificial Intelligence Laboratory
dc.subject	AI
dc.subject	optimal rates
dc.subject	reproducing kernel Hilbert space
dc.subject	effective dimension
dc.title	Risk Bounds for Regularized Least-squares Algorithm with Operator-valued kernels