Stability of Rotor Hopfield Neural Networks With Synchronous Mode

被引：24

作者：

Kobayashi, Masaki ^{[1
]}

机构：

[1] Univ Yamanashi, Ctr Math Sci, Kofu, Yamanashi 4008511, Japan

来源：

IEEE TRANSACTIONS ON NEURAL NETWORKS AND LEARNING SYSTEMS | 2018年 / 29卷 / 03期

关键词：

Complex-valued neural networks; projection rule; rotor Hopfield neural networks (RHNNs); stability condition; synchronous mode; VALUED ASSOCIATIVE MEMORY; PROJECTION RULE; RELAXATION;

D O I：

10.1109/TNNLS.2016.2635140

中图分类号：

TP18 [人工智能理论];

学科分类号：

081104 ; 0812 ; 0835 ; 1405 ;

摘要：

A complex-valued Hopfield neural network (CHNN) is a model of a Hopfield neural network using multistate neurons. The stability conditions of CHNNs have been widely studied. A CHNN with a synchronous mode will converge to a fixed point or a cycle of length 2. A rotor Hopfield neural network (RHNN) is also a model of a multistate Hopfield neural network. RHNNs have much higher storage capacity and noise tolerance than CHNNs. We extend the theories regarding the stability of CHNNs to RHNNs. In addition, we investigate the stability of RHNNs with the projection rule. Although a CHNN with projection rule can be trapped at a cycle, an RHNN with projection rule converges to a fixed point. This is one of the great advantages of RHNNs.

引用

页码：744 / 748

页数：5

共 36 条

[1] [Anonymous], 2003, SERIES INNOVATIVE IN
[2] Aoki H, 2002, ICONIP'02: PROCEEDINGS OF THE 9TH INTERNATIONAL CONFERENCE ON NEURAL INFORMATION PROCESSING, P1084
[3] Aoki H, 2000, INT C PATT RECOG, P626, DOI 10.1109/ICPR.2000.906153
[4] Arena P., 1998, LECT NOTES CONTROL I
[5] Hirose A., 2013, Complex-Valued Neural Networks
[6] Hirose A., SERIES STUDIES COMPU
[7] Isokawa T., 2010, P IEEE WORLD C COMP, P1281
[8] Complex-valued multistate neural associative memory
Jankowski, S
Lozowski, A
Zurada, JM
[J]. IEEE TRANSACTIONS ON NEURAL NETWORKS, 1996, 7 (06): : 1491 - 1496
[9] Kitahara Michimasa, 2010, 2010 IEEE/ACIS 9th International Conference on Computer and Information Science (ICIS 2010), P497, DOI 10.1109/ICIS.2010.38
[10] Kitahara M., 2009, P INT S NONL THEOR I, P399

← 1 2 3 4 →