Dynamic behaviors in chaotic bidirectional associative memory

Authors:
Aluizio F. R. Araú/jo;Luciana P. P. Bueno;Marcí/lia A. Campos
Affiliations:
(Correspd. Tel.: +55 81 2126-8430/ aluizioa@cin.ufpe.br) Center of Informatics, Federal University of Pernambuco - UFPE, Av. Prof. Luí/s Freire, s/n. Cidade Universitá/ria. CEP 50740-540, ...;Departament of Electrical Engineering, University of Sã/o Paulo - USP, Av. Trabalhador Sã/o Carlense, 400. CEP 13566-590, Sã/o Carlos, SP, Brazil;Center of Informatics, Federal University of Pernambuco - UFPE
Venue:
Journal of Intelligent & Fuzzy Systems: Applications in Engineering and Technology - VIII Brazilian Symposium on Neural Networks
Year:
2007

Citing 5
Cited 1

Bidirectional associative memories

IEEE Transactions on Systems, Man and Cybernetics
Practical numerical algorithms for chaotic systems

Practical numerical algorithms for chaotic systems
Linear system theory (2nd ed.)

Linear system theory (2nd ed.)
Associative dynamics in a chaotic neural network

Neural Networks
Testing chaotic dynamics via Lyapunov exponents

Physica D

Logistic Neural Networks: Their chaotic and pattern recognition properties

Neurocomputing

Quantified Score

Hi-index	0.00

Visualization

Abstract

A family of chaotic bidirectional associative memory models (C-BAM family) is introduced through the inclusion of the chaotic neuron in the original BAM family. Each parameter in the chaotic neuron has its influence upon the distinct dynamics estimated throughout experimental design. Based on this analysis, values of parameters are set to illustrate the existence of behaviors such as bifurcation, deterministic chaos and crisis. The existence of the chaotic dynamics is confirmed by calculation of Lyapunov exponents. C-BAM family also presented diversity of patterns recalled, including the occurrence of excursions over all stored memories for some sets of parameters. Hence, C-BAM family can access patterns that original BAM family cannot.