Solving ability of Hopfield neural network with chaotic noise and burst noise for quadratic assignment problem
Yoko Uwate, Yoshifumi Nishio, Tetsushi Ueta, Tohru Kawabe, Tohru Ikeguchi · 2003
Solving combinatorial optimization problems is one of the important applications of the neural network. Many researchers have reported that exploiting chaos achieves good solving ability. However, the reason for the good effect of chaos has not been clarified yet. In this article, intermittent chaos noise near three-periodic window and burst noise generated by the Gilbert model are applied to the Hopfield neural network for quadratic assignment problem. By computer simulations we confirm that the burst noise generated by the Gilbert model is effective to solve the quadratic assignment problem and we can say that the existence of the laminar part and the burst part is one reason of the good performance of the Hopfield NN with chaos noise.