EVOLVING SPACE-FILLING CURVES TO DISTRIBUTE RADIAL BASIS FUNCTIONS OVER AN INPUT SPACE

被引:75
作者
WHITEHEAD, BA [1 ]
CHOATE, TD [1 ]
机构
[1] ERC INC,TULLAHOMA,TN 37388
来源
IEEE TRANSACTIONS ON NEURAL NETWORKS | 1994年 / 5卷 / 01期
关键词
D O I
10.1109/72.265957
中图分类号
TP18 [人工智能理论];
学科分类号
081104 ; 0812 ; 0835 ; 1405 ;
摘要
An evolutionary neural network training algorithm is proposed for radial basis function (RBF) networks. The locations of basis function centers are not directly encoded in a genetic string, but are governed by space-filling curves whose parameters evolve genetically. This encoding causes each group of co-determined basis functions to evolve to fit a region of the input space. A network produced from this encoding is evaluated by training its output connections only. Networks produced by this evolutionary algorithm appear to have better generalization performance on the Mackey-Glass time series than corresponding networks whose centers are determined by k-means clustering.
引用
收藏
页码:15 / 23
页数:9
相关论文
共 44 条
[1]  
BARUN H, 1993, 2ND P ANN C EV PROGR, P62
[2]   Improving the Generalization Properties of Radial Basis Function Neural Networks [J].
Bishop, Chris .
NEURAL COMPUTATION, 1991, 3 (04) :579-588
[3]   GENERAL ASYMMETRIC NEURAL NETWORKS AND STRUCTURE DESIGN BY GENETIC ALGORITHMS [J].
BORNHOLDT, S ;
GRAUDENZ, D .
NEURAL NETWORKS, 1992, 5 (02) :327-334
[4]  
Broomhead D. S., 1988, Complex Systems, V2, P321
[5]  
CAUDELL TP, 1989, 3RD P INT C GEN ALG, P370
[6]   ORTHOGONAL LEAST-SQUARES LEARNING ALGORITHM FOR RADIAL BASIS FUNCTION NETWORKS [J].
CHEN, S ;
COWAN, CFN ;
GRANT, PM .
IEEE TRANSACTIONS ON NEURAL NETWORKS, 1991, 2 (02) :302-309
[7]  
de Garis H., 1990, Machine Learning: Proceedings of the Seventh International Conference (1990), P132
[8]  
DOMINIC S, 1991, IJCNN 91 SEATTLE
[9]   PREDICTING CHAOTIC TIME-SERIES [J].
FARMER, JD ;
SIDOROWICH, JJ .
PHYSICAL REVIEW LETTERS, 1987, 59 (08) :845-848
[10]   EVOLVING NEURAL NETWORKS [J].
FOGEL, DB ;
FOGEL, LJ ;
PORTO, VW .
BIOLOGICAL CYBERNETICS, 1990, 63 (06) :487-493