A New Strategy for Designing Bidirectional Associative Memories

A comprehensive strategy for bidirectional associative memories (BAMs) is presented, which enhances storage capacity greatly. The design strategy combines the dummy augmentation encoding method with optimal gradient descent algorithm. The proposed method increases the storage capacity performance of BAM to its upper limit compared with original Kosko method and optimal gradient descent algorithm. Computer simulations and comparison are given based on three methods to demonstrate the performance improvement of the proposed strategy.

Authors and Affiliations

1. Gengsheng Zheng, School of Computer Science and Technology, Wuhan University, Wuhan, Hubei, 430072, China

   Gengsheng Zheng

2. Sidney Nascimento Givigi Junior, Department of Systems and Computer Engineering, Carleton University, Ottawa, ON, K1S 5B6, Canada

   Sidney Nascimento Givigi

3. Weiyu Zheng, Department of Systems and Computer Engineering, Carleton University, Ottawa, ON, K1S 5B6, Canada

   Weiyu Zheng

Editors and Affiliations

1. Department of Mechanical and Automation Engineering, The Chinese University of Hong Kong, Shatin, New Territories, Hong Kong, China

   Jun Wang

2. The Key Laboratory of Optoelectric Technology & Systems, Ministry of Education, China

   Xiaofeng Liao

3. Computational Intelligence Laboratory, School of Computer Science and Engineering, University of Electronic Science and Technology of China, 610054, Chengdu, P.R. China

   Zhang Yi

Reprints and permissions

© 2005 Springer-Verlag Berlin Heidelberg

Policies and ethics