Modifications to the sliding-window kernel RLS algorithm for time-varying nonlinear systems: Online resizing of the kernel matrix
A kernel-based recursive least-squares algorithm that implements a fixed size ldquosliding-windowrdquo technique has been recently proposed for fast adaptive nonlinear filtering applications. We propose a methodology of resizing the kernel matrix to assist in system identification of time-varying no...
| Main Author: | |
|---|---|
| Other Authors: | |
| Format: | Article |
| Language: | en_US |
| Published: |
Institute of Electrical and Electronics Engineers
2010
|
| Subjects: | |
| Online Access: | http://hdl.handle.net/1721.1/59418 |
| _version_ | 1826213216362954752 |
|---|---|
| author | Julian, Brian John |
| author2 | Massachusetts Institute of Technology. Computer Science and Artificial Intelligence Laboratory |
| author_facet | Massachusetts Institute of Technology. Computer Science and Artificial Intelligence Laboratory Julian, Brian John |
| author_sort | Julian, Brian John |
| collection | MIT |
| description | A kernel-based recursive least-squares algorithm that implements a fixed size ldquosliding-windowrdquo technique has been recently proposed for fast adaptive nonlinear filtering applications. We propose a methodology of resizing the kernel matrix to assist in system identification of time-varying nonlinear systems. To be applicable in practice, the modified algorithm must preserve its ability to operate online. Given a bound on the maximum kernel matrix size, we define the set of all obtainable sizes as the resizing range. We then propose a simple online technique that resizes the kernel matrix within the resizing range. The modified algorithm is applied to the nonlinear system identification problem that was used to evaluate the original algorithm. Results show that an increase in performance is achieved without increasing the original algorithm's computation time. |
| first_indexed | 2024-09-23T15:45:18Z |
| format | Article |
| id | mit-1721.1/59418 |
| institution | Massachusetts Institute of Technology |
| language | en_US |
| last_indexed | 2024-09-23T15:45:18Z |
| publishDate | 2010 |
| publisher | Institute of Electrical and Electronics Engineers |
| record_format | dspace |
| spelling | mit-1721.1/594182022-09-29T15:56:24Z Modifications to the sliding-window kernel RLS algorithm for time-varying nonlinear systems: Online resizing of the kernel matrix Julian, Brian John Massachusetts Institute of Technology. Computer Science and Artificial Intelligence Laboratory Massachusetts Institute of Technology. Department of Electrical Engineering and Computer Science Julian, Brian John Julian, Brian John time-varying filters nonlinear filters least squares methods learning systems identification A kernel-based recursive least-squares algorithm that implements a fixed size ldquosliding-windowrdquo technique has been recently proposed for fast adaptive nonlinear filtering applications. We propose a methodology of resizing the kernel matrix to assist in system identification of time-varying nonlinear systems. To be applicable in practice, the modified algorithm must preserve its ability to operate online. Given a bound on the maximum kernel matrix size, we define the set of all obtainable sizes as the resizing range. We then propose a simple online technique that resizes the kernel matrix within the resizing range. The modified algorithm is applied to the nonlinear system identification problem that was used to evaluate the original algorithm. Results show that an increase in performance is achieved without increasing the original algorithm's computation time. 2010-10-20T12:40:33Z 2010-10-20T12:40:33Z 2009-05 2009-04 Article http://purl.org/eprint/type/ConferencePaper 978-1-4244-2353-8 1520-6149 INSPEC Accession Number: 10701149 http://hdl.handle.net/1721.1/59418 en_US http://dx.doi.org/10.1109/ICASSP.2009.4960352 IEEE International Conference on Acoustics, Speech and Signal Processing, 2009. ICASSP 2009. Article is made available in accordance with the publisher's policy and may be subject to US copyright law. Please refer to the publisher's site for terms of use. application/pdf Institute of Electrical and Electronics Engineers IEEE |
| spellingShingle | time-varying filters nonlinear filters least squares methods learning systems identification Julian, Brian John Modifications to the sliding-window kernel RLS algorithm for time-varying nonlinear systems: Online resizing of the kernel matrix |
| title | Modifications to the sliding-window kernel RLS algorithm for time-varying nonlinear systems: Online resizing of the kernel matrix |
| title_full | Modifications to the sliding-window kernel RLS algorithm for time-varying nonlinear systems: Online resizing of the kernel matrix |
| title_fullStr | Modifications to the sliding-window kernel RLS algorithm for time-varying nonlinear systems: Online resizing of the kernel matrix |
| title_full_unstemmed | Modifications to the sliding-window kernel RLS algorithm for time-varying nonlinear systems: Online resizing of the kernel matrix |
| title_short | Modifications to the sliding-window kernel RLS algorithm for time-varying nonlinear systems: Online resizing of the kernel matrix |
| title_sort | modifications to the sliding window kernel rls algorithm for time varying nonlinear systems online resizing of the kernel matrix |
| topic | time-varying filters nonlinear filters least squares methods learning systems identification |
| url | http://hdl.handle.net/1721.1/59418 |
| work_keys_str_mv | AT julianbrianjohn modificationstotheslidingwindowkernelrlsalgorithmfortimevaryingnonlinearsystemsonlineresizingofthekernelmatrix |