Structure adaptation of polynomial stochastic neural nets using learning automata technique

The paper is concerned with the selection of a number of nodes in polynomial artificial neural nets containing stochastic noise perturbations in the outputs of each node. The suggested approach is based on a reinforcement learning technique. To solve this optimization problem we introduce a special performance index in such a way that the best number of nodes corresponds to the minimum point of the suggested criterion. This criterion presents a linear combination of a residual minimization functional and some "generalized variance" of the involved disturbances of random nature. A large value of the noise variance leads to a different optimal number of neurons in a neural network because of the "interference" effect. Simulation modeling results are presented to illustrate the effectiveness of the suggested approach.