Гібридна система обчислювального інтелекту, що самонавчається, на основі спайк-нейронної мережі

Abstract

The self-learning hybrid systems of calculable intellect are considered on the basis of spiking neural network. Generalized architecture of such system that allows of hierarchical fuzzy clustering is proposed. Based on the generalized Hebb rule of self-learning, a learning algorithm of the hybrid system is proposed. Synthesized architecture of the hybrid system is stated in terms of the Laplace transform.