Sigmoids Distinguish More Efficiently Than Heavisides

被引：18

作者：

Sontag, Eduardo D. ^{[1
]}

机构：

[1] Rutgers Univ, SYCON Rutgers Ctr Syst & Control, Dept Math, New Brunswick, NJ 08903 USA

来源：

NEURAL COMPUTATION | 1989年 / 1卷 / 04期

关键词：

D O I：

10.1162/neco.1989.1.4.470

中图分类号：

TP18 [人工智能理论];

学科分类号：

081104 ; 0812 ; 0835 ; 1405 ;

摘要：

Every dichotomy on a 2k-point set in R(N )can be implemented by a neural net with a single hidden layer containing k sigmoidal neurons. If the neurons were of a hardlimiter (Heaviside) type, 2k - 1 would be in general needed.

引用

页码：470 / 472

页数：3