DocumentCode
1898457
Title
Backtracking orthogonal least squares algorithm for model selection
Author
Chng, Eng Siong ; Mulgrew, Bernard
Author_Institution
Dept. of Electr. Eng., Edinburgh Univ.
fYear
1994
fDate
34375
Firstpage
42644
Lastpage
42649
Abstract
The orthogonal least squares (OLS) algorithm is an efficient implementation of the forward-selection method for subset model selection. The ability to find good subset parameters with only a linearly increasing computational requirement makes this method attractive for practical implementations. This paper examines why forward-selection technique can fail to find optimum subset models and presents a modification scheme to improve the selection process
Keywords
computational complexity; least squares approximations; modelling; signal processing; backtracking orthogonal least squares algorithm; computational requirement; forward-selection method; modification scheme; optimum subset models; subset model selection;
fLanguage
English
Publisher
iet
Conference_Titel
Mathematical Aspects of Digital Signal Processing, IEE Colloquium on
Conference_Location
London
Type
conf
Filename
297468
Link To Document