Convergence analysis of continuous-time neural networks

Abstract

The energy function of continuous-time neural network has been analyzed for testing the existence of stationary points and the global convergence of network. The energy function always has only one stationary point which is a saddle point in the unconstrained space when the total conductance of neuron's input is zero (G1 = 0). However, the stationary points exist only inside the hy-percube Rnε[0,1] when the total conductance of neuron's input is not zero (G, ≠ 0). The Hessian matrix of the energy function is used for testing the global convergence of the network. © Springer-Verlag Berlin Heidelberg 2006.