Partially linear models and least squares support vector machines

Within the context of nonlinear system identification, the LS-SVM formulation is extended to define a partially linear LS-SVM in order to identify a model containing a linear part and a nonlinear component. For a given kernel, a unique solution exists when the parametric part has full column rank, a...

Full description

Saved in:
Bibliographic Details
Main Authors: Espinoza, M., Suykens, J.A.K., De Moor, B.
Format: Conference Proceeding
Language:English
Subjects:
Applied sciences
Computer science
control theory
systems
Context modeling
Control theory. Systems
Cost function
Exact sciences and technology
Kernel
Least squares approximation
Least squares methods
Linear systems
Modelling and identification
Nonlinear systems
Predictive models
Quadratic programming
Support vector machines
Online Access:Request full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:Within the context of nonlinear system identification, the LS-SVM formulation is extended to define a partially linear LS-SVM in order to identify a model containing a linear part and a nonlinear component. For a given kernel, a unique solution exists when the parametric part has full column rank, although identifiability problems can arise for certain structures. The solution has close links with traditional semiparametric techniques from the statistical literature. The properties of the model are illustrated by Monte Carlo simulations over different structures, and iterative forecasting examples for Hammerstein and other systems show a good global performance and an accurate identification of the linear part.
ISSN:0191-2216
DOI:10.1109/CDC.2004.1429230