APPROXIMATION-THEORY AND FEEDFORWARD NETWORKS

被引：203
作者：
BLUM, EK
LI, LK
机构：
来源：
NEURAL NETWORKS | 1991年 / 4卷 / 04期
关键词：
FEEDFORWARD NETWORKS; APPROXIMATION; 2-LAYER NETS; 3-LAYER NETS;
D O I：
10.1016/0893-6080(91)90047-9
中图分类号：
TP18 [人工智能理论];
学科分类号：
081104 ; 0812 ; 0835 ; 1405 ;
摘要：
Approximation of real functions by feedforward networks of the usual kind is shown to be based on the fundamental principle of approximation by piecewise-constant functions. This principle underlies a simple construction given for three-layer networks and suggests possible difficulties in determining two-layer networks.
引用
收藏
页码：511 / 515
页数：5
相关论文
共 12 条
[1]  
BLUM EK, 1972, NUMERICAL ANAL COMPU
[2]  
Cybenko G., 1989, Mathematics of Control, Signals, and Systems, V2, P303, DOI 10.1007/BF02551274
[3]  
DEFIGUEIREDO RJP, 1980, IEEE T AUTOMATIC CON, V25, P506
[4]   ON THE APPROXIMATE REALIZATION OF CONTINUOUS-MAPPINGS BY NEURAL NETWORKS [J].
FUNAHASHI, K .
NEURAL NETWORKS, 1989, 2 (03) :183-192
[5]   MULTILAYER FEEDFORWARD NETWORKS ARE UNIVERSAL APPROXIMATORS [J].
HORNIK, K ;
STINCHCOMBE, M ;
WHITE, H .
NEURAL NETWORKS, 1989, 2 (05) :359-366
[6]  
Kolmogorov A. N., 1961, AM MATH SOC TRANSL, V17, P369
[7]  
KOLMOGOROV AN, 1956, DOKL AKAD NAUK SSSR+, V108, P179
[8]  
LAPEDES A, 1987, NEURAL INFORMATION P, P442, DOI DOI 10.1142/9789814434102_0012
[9]  
LI LK, 1990, LEARNING APPLICATION
[10]  
Lippman R. P., 1987, IEEE ASSP MAGAZI APR, P4
12