Analysis and Design of Associative Memories Based on Recurrent Neural Networks with Linear Saturation Activation Functions and Time-Varying Delays

Authors:
Zhigang Zeng;Jun Wang
Affiliations:
School of Automation, Wuhan University of Technology, Wuhan, Hubei, 430070, China, and Department of Mechanical and Automation Engineering, Chinese University of Hong Kong, Shatin, New Territories ...;Department of Mechanical and Automation Engineering, Chinese University of Hong Kong, Shatin, New Territories, Hong Kong jwang@mae.cuhk.edu.hk
Venue:
Neural Computation
Year:
2007

Citing 5
Cited 14

On the design of BSB neural associative memories using semidefinite programming

Neural Computation
Multistability analysis for recurrent neural networks with unsaturating piecewise linear transfer functions

Neural Computation
Multiperiodicity and Exponential Attractivity Evoked by Periodic External Inputs in Delayed Cellular Neural Networks

Neural Computation
An Optimization Approach to Design of Generalized BSB Neural Associative Memories

Neural Computation
A new synthesis approach for feedback neural networks based on the perceptron training algorithm

IEEE Transactions on Neural Networks

Disturbance Attenuating Controller Design for a Class of Nonlinear Systems with Unknown Time-Delay

ISNN '08 Proceedings of the 5th international symposium on Neural Networks: Advances in Neural Networks
Passivity Analysis of a General Form of Recurrent Neural Network with Multiple Delays

ISNN 2009 Proceedings of the 6th International Symposium on Neural Networks: Advances in Neural Networks - Part III
2009 Special Issue: Associative memories based on continuous-time cellular neural networks designed using space-invariant cloning templates

Neural Networks
Delayed neural networks with multistable almost periodic solutions

IJCNN'09 Proceedings of the 2009 international joint conference on Neural Networks
Analysis and design of associative memories based on cellular neural networks with space-invariant cloning templates

IJCNN'09 Proceedings of the 2009 international joint conference on Neural Networks
Analysis and synthesis of associative memories based on brain-state-in-a-box neural networks

IJCNN'09 Proceedings of the 2009 international joint conference on Neural Networks
Multistability and new attraction basins of almost-periodic solutions of delayed neural networks

IEEE Transactions on Neural Networks
A gaussian attractor network for memory and recognition with experience-dependent learning

Neural Computation
Synchronization control of a class of memristor-based recurrent neural networks

Information Sciences: an International Journal
Analysis and design of associative memories based on recurrent neural network with discontinuous activation functions

Neurocomputing
Guaranteed cost synchronization of chaotic cellular neural networks with time-varying delay

Neural Computation
Dynamical stability analysis of multiple equilibrium points in time-varying delayed recurrent neural networks with discontinuous activation functions

Neurocomputing
A neural network with a single recurrent unit for associative memories based on linear optimization

Neurocomputing
A mode-dependent approach to state estimation of recurrent neural networks with Markovian jumping parameters and mixed delays

Neural Networks

Quantified Score

Hi-index	0.00

Visualization

Abstract

In this letter, some sufficient conditions are obtained to guarantee recurrent neural networks with linear saturation activation functions, and time-varying delays have multiequilibria located in the saturation region and the boundaries of the saturation region. These results on pattern characterization are used to analyze and design autoassociative memories, which are directly based on the parameters of the neural networks. Moreover, a formula for the numbers of spurious equilibria is also derived. Four design procedures for recurrent neural networks with linear saturation activation functions and time-varying delays are developed based on stability results. Two of these procedures allow the neural network to be capable of learning and forgetting. Finally, simulation results demonstrate the validity and characteristics of the proposed approach.