Industrial Automatic Control Systems and Controllers Annotation << Back Perceptron Structural Optimization Using the Apparatus of Evolutionary Modeling for the Task of Predicting the Values of Time Series V.S. Tormozov, A.L. Zolkin, K.A. Vasilenko The article proposes a method that allows using the stochastic mutation function with a constant coefficient of variation to solve the complex combinatorial problem of structural optimization of an artificial neural network with a large dimension of the space of optimization parameters. A study is given to assess the applicability of this approach to solve the problem. This method is proposed to be used for the task of forecasting the values of time series. Keywords: artificial neural network; genetic algorithm; forecasting; time series; structural optimization. DOI: 10.25791/asu.8.2020.1210 Pp. 40-45.