MATLAB PROGRAM CODES FOR BIDIRECTIONAL ASSOCIATIVE MEMORY NETWORKS
Keywords:Artificial neural networks, bidirectional associative memory, auto-association, hetero-association
Objective Neural networks are being used for solving problems in various diverse areas including education, research, business, management, and many more. In this article, models describing the dynamics of bidirectional associative memory (BAM) neural networks are considered.
Methods: MATLAB, the numerical computing environment and programming language is used for solving certain problems associated with BAM.
Results: The concept of BAM networks is improved so that it can be applied to a wider class of networks. Algorithm for solving BAM problems is studied. And also the MATLAB program codes to find the weight matrix, to test the net with input, and to generate activation functions are accompanied.
Conclusion: MATLAB programming can be effectively used to solve the problems associated with BAM.
2. Bart Kosko. Bidirectional Associative Memories. IEEE Transactions on Systems, Man, and Cybernetics. 1988; 18:49-60.
3. Chartier S, Boukadoum M. Encoding static and temporal patterns with a bidirectional heteroassociative memory. J of Applied Mathematics. 2011; 68:1-34.
4. Govindan V K, Sivaprasad A P. Character Recognition- A review. Pattern recognition 1999; 23:671-83.
5. Haikin S. Neural Networks: A comprehensive Foundation. New York: Macmillan College Publishing Company; 2014.
6. Zurada J M. Introduction to Artificial Neural Systems with Applications. Mumbai, India: Jaico Publication House; 2004.
7. Xu Z B, Leung Y, He X W. Asymmetrical Bidirectional Associative Memories. IEEE Transactions on Systems, Man and Cybernetics. 2008; 38: 1558-64.
8. Kalyan V A. On the recent developments and programming of bidirectional associative memory network. Int Research J Printing Area. 2018;68:152-61