Solving Ability of Hopfield Neural Network with Mix Noise for Quadratic Assignment Problem
Yoshifumi Tada, Yoko Uwate, Yoshifumi Nishio · 2006
Solving combinatorial optimization problem is one of the im-portant applications of neural network (abbr. NN). However, the solutions are often trapped into a local minimum and do not reach the global minimum. In order to avoid this critical problem, several people proposed the method adding some kinds of noise. In this study, we consider torus noise generated by the sine circle map for the Hopfield NN. By computer simula-tions, solving abilities of Hopfield NN for quadratic assign-ment problem (QAP) with various kinds of noises based on the torus noise are investigated. 1.