IMPROVEMENT OF THE HOPFIELD MODEL WITH INTERPATTERN ASSOCIATION
IPA模型对Hopfield模型神经网络的改进

The interpattern association is proposed to suppress the stable false and oscillatory states in neural networks based on the Hopfield model, in consideration of the existence of Boolean logic relations in those states. The digital simulation of neural networks of the Hopfield model and the Hopfield model with interpattern association is shown. The Hopfield model with interpattern association is of higher performance than the Hopfield model.