文章基本信息

标题：Neural Circuits for Pattern Recognition with Small Total Wire Length
本地全文：下载
作者：Robert Albin Legenstein ; Wolfgang Maass
期刊名称：Electronic Colloquium on Computational Complexity
印刷版ISSN：1433-8092
出版年度：2001
卷号：2001
出版社：Universität Trier, Lehrstuhl für Theoretische Computer-Forschung
摘要：One of the most basic pattern recognition problems is whether a certain local feature occurs in some linear array to the left of some other local feature. We construct in this article circuits that solve this problem with an asymptotically optimal number of threshold gates. Furthermore it is shown that much fewer threshold gates are needed if one employs in addition a small number of winner-take-all gates. In either case the circuits that are constructed have linear or almost linear total wire length, and are therefore not unrealistic from the point of view of physical implementations.
关键词：circuit complexity , Neural Networks , pattern recognition , total wire length , winner take all