Orthogonal least squares regression with tunable kernels

Abstract

A novel technique is proposed to construct sparse regression models based on the orthogonal least squares method with tunable kernels. The proposed technique tunes the centre vector and diagonal covariance matrix of individual regressors by incrementally minimising the training mean square error using a guided random search algorithm, and it offers a state-of-the-art method for constructing very sparse models that generalise well.

Publication DOI: https://doi.org/10.1049/el:20050265
Additional Information: © 2005 IEEE. Personal use of this material is permitted. Permission from IEEE must be obtained for all other uses, in any current or future media, including reprinting/republishing this material for advertising or promotional purposes, creating new collective works, for resale or redistribution to servers or lists, or reuse of any copyrighted component of this work in other works.
Uncontrolled Keywords: Electrical and Electronic Engineering
Publication ISSN: 1350-911X
Last Modified: 28 Feb 2024 08:11
Date Deposited: 06 Nov 2019 16:01
Full Text Link:
Related URLs: http://www.scop ... tnerID=8YFLogxK (Scopus URL)
PURE Output Type: Article
Published Date: 2005-04-14
Authors: Chen, S.
Wang, X. X.
Brown, D. J.

Download

[img]

Version: Accepted Version

| Preview

Export / Share Citation


Statistics

Additional statistics for this record