Title :
Training RBF networks with perturbation methods
Author_Institution :
Dept. of Electron. Syst. Eng., Essex Univ., Colchester, UK
fDate :
30 Apr-3 May 1995
Abstract :
This paper describes a gradient descent technique for training radial basis function (RBF) networks which is suitable for hardware implementation. The method dynamically adjusts the positions and the widths of the basis functions so as to reduce the total output error of the network while the output connection weights are being trained. The algorithm is demonstrated by using it to train an RBF network to perform simple logical functions
Keywords :
functions; learning (artificial intelligence); neural nets; perturbation techniques; RBF networks; dynamic adjustment; gradient descent technique; network training; output connection weights; perturbation method; radial basis function; supervised learning; total output error reduction; Clustering algorithms; Equations; Function approximation; Hardware; Neural networks; Perturbation methods; Radial basis function networks; Systems engineering and theory; Training data; Very large scale integration;
Conference_Titel :
Circuits and Systems, 1995. ISCAS '95., 1995 IEEE International Symposium on
Conference_Location :
Seattle, WA
Print_ISBN :
0-7803-2570-2
DOI :
10.1109/ISCAS.1995.523739
