Identification of non linear MISO process using RKHS and Volterra models

Authors:
Okba Taouali;Nabiha Saidi;Hassani Messaoud
Affiliations:
Unité de Recherche d'Automatique, Traitement de Signal et Image, Ecole Nationale d'Ingénieur Monastir, Monastir, Tunisia;Unité de Recherche d'Automatique, Traitement de Signal et Image, Ecole Nationale d'Ingénieur Monastir, Monastir, Tunisia;Unité de Recherche d'Automatique, Traitement de Signal et Image, Ecole Nationale d'Ingénieur Monastir, Monastir, Tunisia
Venue:
WSEAS TRANSACTIONS on SYSTEMS
Year:
2009

Citing 5
Cited 3

The nature of statistical learning theory

The nature of statistical learning theory
Nonlinear component analysis as a kernel eigenvalue problem

Neural Computation
Kernel partial least squares regression in reproducing kernel hilbert space

The Journal of Machine Learning Research
Efficient implementation and performance evaluation of the second order volterra filter based on the MMD approximation

WSEAS Transactions on Circuits and Systems
Induction machine fault detection using support vector machine based classifier

WSEAS TRANSACTIONS on SYSTEMS

Weight-decay regularization in reproducing Kernel Hilbert spaces by variable-basis schemes

WSEAS Transactions on Mathematics
Supervised learning with kernel methods

WAMUS'10 Proceedings of the 10th WSEAS international conference on Wavelet analysis and multirate systems
A new approach for identification of MIMO non linear system with RKHS model

WSEAS Transactions on Information Science and Applications

Quantified Score

Hi-index	0.00

Visualization

Abstract

This paper treats the comparison between the Volterra model and Reproducing Kernel Hilbert Space (RKHS) model in Multiple Input Single Output (MISO) case. The RKHS model uses the Statistical learning theory to find a solution of a regularization risk. It is characterise by a linear combination of the kernels function. The complexity of Volterra model is depending of the degree and the memory of the model contrarily of the RKHS model which depend only of the number of observations. The performances of both models are evaluated first by using Monte Carlo numerical simulations and then have been tested for modelling of a chemical reactor and results are successful.