Analysis and synthesis of neural networks with lower block triangular interconnecting structure

摘要

The authors present a qualitative analysis of Hopfield-type neural network models with lower block triangular interconnecting structure. Such networks are viewed as interconnected systems and the present results are phrased in terms of the qualitative properties of the subsystems of the networks and in terms of the properties of the interconnecting structure of the networks. The results address the stability properties of equilibrium points and estimates of trajectory properties. These results enable one to devise a design method for the class of neural networks considered to establish desired relationships between scalar-valued analog input signals and output signals in binary form.