Chaotic neural networks with reinforced self-feedbacks and its application to N-Queen problem Original Research Article

The chaotic neural network (CNN) has a characteristic of escaping from a local minimum of the energy function, so that it can find a global minimum more easily as compared with the Hopfield’s model. However, it is sometimes difficult to escape from the local minimum by only the chaotic behavior. To overcome it, the CNN with reinforced self-feedbacks is proposed in this paper. The proposed algorithm gradually intensifies the self-feedback connection of the active neurons and attempts to escape from the local minimum. In order to confirm the effectiveness, it is applied to the N-Queen problem, N= 50–1000. From experimental results, the average of success rate of obtaining a solution is improved from 30 to 90% in N=1000.