A GENERAL-SOLUTION FOR COMBINATORIAL PROBLEMS BY NEURAL NETWORKS

Yao et al. have proposed the Dynamic-Static Neural network (DS-net) wh ich includes the Tank and Hopfield model for linear programming proble ms as its special case. This paper proposes a modified DS-net which is capable of solving optimization problems with quadratic constraints. Because the 3-satisfiability problem of Boolean formula can be transfo rmed to problems of this type, the model is applied to the problem. Th e simulation experiments using a digital computer show that this formu lation bring the merit of greedy methods to the neural network solutio n and the initial state satisfying constraints is excellent for obtain ing the optimal solution.