Architecture of neural network complex for forecasting and analysis of time series based on the neural network spectral analysis

Architecture of neural network complex for forecasting and analysis of time series based on the neural network spectral analysis has been developed. Advantages of neural spectral analysis compared with the existing methods of singular spectral analysis have been shown. Approach to trend and other components forecasting of the time series has been proposed. Architecture improvement option has been presented. The developed neural network complex is advisable to apply for the analysis and forecasting of non-stationary processes in complex dynamic systems.