Decision making model based on neural network with diagonalized synaptic connections

Abstract

In this paper, we propose a decision-making model based on the architecture of a three-layer perceptron with diagonal weighted synaptic connections between the neurons of the input, the latent and the original layers. The evolution of the model is carried out as a task of adaptation of the neural network, which consists of procedures for correction of the number of synaptic connections between the neurons of the input hidden and output layers due to the diagonalization of the matrices of synaptic connections in the basis of the input vector vectors. It is shown that the time of decision making in the diagonalized three-layer neural network is smaller in comparison with the time in the non-diagonalized.