OPTICAL IMPLEMENTATION OF A 2ND-ORDER NEURAL NETWORK DISCRIMINATOR MODEL

被引：5

作者：

HORAN, P ^{[1
]}

UECKER, D ^{[1
]}

ARIMOTO, A ^{[1
]}

机构：

[1] HITACHI LTD, CENT RES LABS, KOKUBUNJI, TOKYO 185, JAPAN

来源：

JAPANESE JOURNAL OF APPLIED PHYSICS PART 2-LETTERS & EXPRESS LETTERS | 1990年 / 29卷 / 07期

关键词：

Higher-order neural nets; Neural networks; Optical neural nets;

D O I：

10.1143/JJAP.29.L1328

中图分类号：

O59 [应用物理学];

学科分类号：

摘要：

Some of the symmetries inherent in a second-order translation invariant classifier model are described. The operational characteristics of this model for pattern recognition tasks such as fault and noise tolerance are explored. A particular optical architecture is detailed which exploits particular features of the model suited to optics and the operation of this system described. © 1990 The Japan Society of Applied Physics.

引用

页码：L1328 / L1331

页数：4

共 7 条

[1] OPTICAL MATRIX MATRIX MULTIPLIER BASED ON OUTER PRODUCT DECOMPOSITION [J].

ATHALE, RA ;

COLLINS, WC .

APPLIED OPTICS, 1982, 21 (12) :2089-2090

[2] LEARNING, INVARIANCE, AND GENERALIZATION IN HIGH-ORDER NEURAL NETWORKS [J].

GILES, CL ;

MAXWELL, T .

APPLIED OPTICS, 1987, 26 (23) :4972-4978

[3]

GRIFFIN RD, 1987, J OPT SOC AM, V4, P118

[4]

MINSKY LML, 1988, PERCEPTRONS

[5]

MINSKY ML, 1988, PERCEPTRONS, pCH6

[6]

REID MB, 1989, P JOINT INT C NEUR N, P689

[7]

RUMELHART D, 1986, PARALLEL DISTRIBUTED, pCH5

← 1 →