STATE OBSERVABILITY IN RECURRENT NEURAL NETWORKS

被引：25

作者：

ALBERTINI, F ^{[1
]}

SONTAG, ED ^{[1
]}

机构：

[1] RUTGERS STATE UNIV,SYCON,RUTGERS CTR SYST & CONTROL,NEW BRUNSWICK,NJ 08903

来源：

关键词：

RECURRENT NEURAL NETWORKS; OBSERVABILITY;

D O I：

10.1016/0167-6911(94)90054-X

中图分类号：

TP [自动化技术、计算机技术];

学科分类号：

0812 ;

摘要：

We obtain a characterization of observability for a class of nonlinear systems which appear in neural networks research.

引用

页码：235 / 244

页数：10

共 7 条

[1]

ALBERTINI F, 1992, PROCEEDINGS OF THE 31ST IEEE CONFERENCE ON DECISION AND CONTROL, VOLS 1-4, P26, DOI 10.1109/CDC.1992.371799

[2]

ALBERTINI F, 1993, IN PRESS ARTIFICIAL

[3]

[Anonymous], 1991, INTRO THEORY NEURAL, DOI DOI 10.1201/9780429499661

[4] UNIQUENESS OF CIRCUITS AND SYSTEMS CONTAINING ONE NONLINEARITY [J].

BOYD, SP ;

CHUA, LO .

[5] STRUCTURE OF A CLASS OF NONLINEAR-SYSTEMS [J].

SMITH, WW ;

RUGH, WJ .

[6]

Sontag E.D., 1990, MATH CONTROL THEORY

SONTAG, ED .

IEEE TRANSACTIONS ON CIRCUITS AND SYSTEMS, 1979, 26 (05) :342-356