GENERAL NEURAL UNIT - RETRIEVAL PERFORMANCE

被引:14
作者
ALEKSANDER, I [1 ]
MORTON, HB [1 ]
机构
[1] BRUNEL UNIV,DEPT HUMAN RESOURCES,UXBRIDGE UB8 3PH,MIDDX,ENGLAND
关键词
NEURAL NETWORKS;
D O I
10.1049/el:19911105
中图分类号
TM [电工技术]; TN [电子技术、通信技术];
学科分类号
0808 ; 0809 ;
摘要
A general neural unit is described which is capable of retrieving learned prototype images at internal connections when stimulated by distorted versions of the prototypes at the external terminals. Theoretical predictions of optimal feedback for best retrieval performance are given with experimental confirmation.
引用
收藏
页码:1776 / 1778
页数:3
相关论文
共 4 条
[1]  
ALEKSANDER I, 1990, COMPUT CONTROL ENG, V1, P259
[2]   PERFECT AUTO-ASSOCIATORS USING RAM-TYPE NODES [J].
LUCY, J .
ELECTRONICS LETTERS, 1991, 27 (10) :799-800
[3]  
NTOURNTOUFIS P, 1991, P IJCNN SEATTLE, P959
[4]  
WONG KYM, 1988, P NEURO 88 PARIS, P424