6533b82efe1ef96bd129337b

RESEARCH PRODUCT

Learning non-linear time-scales with kernel -filters

Jordi Munoz-mariJesus Requena-carrionJosé Luis Rojo-álvarezManel Martínez-ramónGustau Camps-valls

subject

TelecomunicacionesSupport vector machinesbusiness.industryCognitive NeuroscienceNonlinear System IdentificationPattern recognitionKernel principal component analysisComputer Science ApplicationsKernel methodMercer's KernelArtificial IntelligenceVariable kernel density estimationString kernelKernel embedding of distributionsPolynomial kernelRadial basis function kernelGamma-FiltersArtificial intelligenceTree kernelbusinessMathematics

description

A family of kernel methods, based on the @c-filter structure, is presented for non-linear system identification and time series prediction. The kernel trick allows us to develop the natural non-linear extension of the (linear) support vector machine (SVM) @c-filter [G. Camps-Valls, M. Martinez-Ramon, J.L. Rojo-Alvarez, E. Soria-Olivas, Robust @c-filter using support vector machines, Neurocomput. J. 62(12) (2004) 493-499.], but this approach yields a rigid system model without non-linear cross relation between time-scales. Several functional analysis properties allow us to develop a full, principled family of kernel @c-filters. The improved performance in several application examples suggests that a more appropriate representation of signal states is achieved.

yearjournalcountryeditionlanguage
2009-01-01Neurocomputing
https://doi.org/10.1016/j.neucom.2008.10.004