DocumentCode :
533153
Title :
A novel OLS algorithm for training RBF neural networks with automatic model selection
Author :
Zhou, Peng ; Li, Dehua ; Wu, Hong ; Chen, Feng
Author_Institution :
Inst. for Pattern Recognition & Artificial Intell., Huazhong Univ. of Sci. & Technol., Wuhan, China
Volume :
10
fYear :
2010
fDate :
22-24 Oct. 2010
Abstract :
Orthogonal Least Squares (OLS) algorithm has been extensively used in basis selection problems for RBF networks, but it is unable perform model selection automatically because the user is required to specify the tolerance?, which is relevant to noises and will be difficult to implement in the real system. therefore, a generic criterion that defines the optimum number of its basis function is proposed. In this paper, Not only is the Bayesian information criteria (BIC) method incorporate into the basis function selection process of the OLS algorithm for assigning its appropriate number, but also we develop a new method to optimize the widths of Gaussian functions in order to improve the generalization performance, The augmented algorithms are employed to the Radial Basis Function Neural Networks (RBFNN) to compare its performance for known and unknown noise nonlinear dynamic systems, Experimental results show the efficacy of this criterion and the importance of a proper choice of basis function widths.
Keywords :
Bayes methods; Gaussian processes; belief networks; generalisation (artificial intelligence); learning (artificial intelligence); least squares approximations; radial basis function networks; Bayesian information criteria; Gaussian functions; OLS algorithm; RBF neural networks training; augmented algorithms; automatic model selection; basis function selection process; generalization performance; nonlinear dynamic systems; orthogonal least squares algorithm; Bayesian methods; Bayesian information criteria; kernel widths; orthogonal least squares; radial basis function networks;
fLanguage :
English
Publisher :
ieee
Conference_Titel :
Computer Application and System Modeling (ICCASM), 2010 International Conference on
Conference_Location :
Taiyuan
Print_ISBN :
978-1-4244-7235-2
Electronic_ISBN :
978-1-4244-7237-6
Type :
conf
DOI :
10.1109/ICCASM.2010.5622900
Filename :
5622900
Link To Document :
بازگشت