Learning from dependent observations

Simple item page
dc.creatorSteinwart, Ingo
dc.creatorHush, Don
dc.creatorScovel, Clint
dc.date2007-07-02
dc.date.accessioned2026-07-07T08:13:42Z
dc.date.available2026-07-07T08:13:42Z
dc.descriptionIn most papers establishing consistency for learning algorithms it is assumed that the observations used for training are realizations of an i.i.d. process. In this paper we go far beyond this classical framework by showing that support vector machines (SVMs) essentially only require that the data-generating process satisfies a certain law of large numbers. We then consider the learnability of SVMs for $\a$-mixing (not necessarily stationary) processes for both classification and regression, where for the latter we explicitly allow unbounded noise.
dc.descriptionsubmitted to Journal of Multivariate Analysis
dc.identifierhttps://arxiv.org/abs/0707.0303
dc.identifierhttp://arxiv.org/abs/0707.0303
dc.identifier.urihttp://salesiana.dossiersoluciones.com/handle/123456789/132860
dc.subjectMachine Learning
dc.subjectMethodology
dc.titleLearning from dependent observations
dc.typetext

Files

Collections

arXiv