A new learning algorithm for radial basis functional network, which is called ABS project learning algorithm, is given and computing results show that it can be used for pattern recognition and other areas.
Powell M J D. Radial basis function for multivariable interpolation review. in algorithms for approximation. Mason J C, Cox M G Ed. Oxford University Press, 1987, 143-167.[2]Powell M J D. Radial basis function in 1990. in advances in numerical analysis Vol.2. Oxford University Press, 1991. 105-201.[3]Broomhead D S, Lowe D. Multivariable function interpolation and daptive networks. Complex System, 1988, No.2: 321-355.[4]Chen S, Cowan C F N, Grant P M. Orthogonal least squares learning algorithm for radial basis function networks. IEEE Trans. on NN, 1991, NN-2: 302-309.[5]Moody J,Darken C. Fast-learning in networks of locally tuned processing units. Neural Comput.,[6]89, 1: 281-294.[7]Poggio T,Girosi F. Networks for approximtion and learning[J].Proc. IEEE.1990, 78:1481-1497[8]Micchelli C A. Interpolation of scatterd data: Distance matrices and conditionally positive definite function[J].Constr. Approx.1986, 2:11-22[9]Chen S, Billings S A. Orthogonal least squares methods and their application to non-linear system identication, International Journal of Control 1989 50: 1873-1896.[10]Abaffy J.[匈], Spedicato E.[意] 著,诸梅芳,宫伴晶译.ABS投影算法一一求解线性和非线性方程组的数学方法.北京:北京工业大学出版社,1991,第三章.