A neural network model based on BP learning with stochastic resonance

We describe the results of computer simulations of the dynamical behavior of a neural network model based on BP learning incorporated stochastic resonance. Such a model can acquire the ability of the avoidance of local minima. Our results show that, under the influence of a weak periodic signal, the network exhibits a maximum in the signal to noise ratio at an optimum noise level: the characteristic signature of stochastic resonance. At this level, the learning abilities of our network indicate a significant improvement compared to the limited abilities of the traditional BP learning neural network.