Bidirectional neural network for clustering problems

There exist several neural networks models for solving NP-hard combinatorial optimization problems. Hopfield networks and self-organizing maps are the two main neural approaches studied. Criticism of these approaches includes the tendency of the Hopfield network to produce infeasible solutions and the lack of generalization of the self-organizing approaches. This paper presents a new bidirectional neural model for solving clustering problems. Bidirectional associative memory (BAM) is the best bidirectional neural architecture known. Typically, this model has ever been used for information storage. In this paper we propose a new neural model with this bidirectional neural architecture for optimization problems, concretely clustering problems. A sample theoretical clustering problem as the p-median problem is used to test the performance of the proposed neural approach against genetic algorithms and traditional heuristic techniques.