Publication: Non-linear System Identification with Composite Relevance Vector Machines
Loading...
Identifiers
Publication date
2007-04
Defense date
Advisors
Tutors
Journal Title
Journal ISSN
Volume Title
Publisher
IEEE
Abstract
Nonlinear system identification based on relevance vector machines (RVMs) has been traditionally addressed by stacking the input and/or output regressors and then performing standard RVM regression. This letter introduces a full family of composite kernels in order to integrate the input and output information in the mapping function efficiently and hence generalize the standard approach. An improved trade-off between accuracy and sparsity is obtained in several benchmark problems. Also, the RVM yields confidence intervals for the predictions, and it is less sensitive to free parameter selection
Description
Keywords
Composite kernels, Nonlinear System Identification, Relevance Vector Machines, RVM
Bibliographic citation
IEEE Signal Processing Letters, Vol. 14, No 4, pp. 279-282, April, 2007