ВЛИЯНИЕ ФИЛЬТРАЦИИ ФУРЬЕ НА ТОЧНОСТЬ ПРОГНОЗА СОЦИАЛЬНО-ЭКОНОМИЧЕСКИХ ПОКАЗАТЕЛЕЙ

Abstract

Discusses the possibility of removing the noise component of the time-series social and economic indicators, using the discrete Fourier transform and applying a window filter. A study of the effectiveness of the method for the generated time series with the additive noise component and without it. The quality of filtering of the time series is determined by the type of the filtering window, which also depends on the accuracy of future forecast values of a time series.