On the global convergence of the gradient neural networks and the attractive region of the equilibrium point
Quanju Zhang · Jounal of Xidian University · 2001
Neural networks for problems of the global optimization in the case of objective function having a finite number of stationary points are proposed. The attractive region of the equilibrium point is given with the global convergence of the equilibrium point set also studied. Our results present a theoretic basis for the active and efficient movement of the neural networks.