The comparative analysis of forecasting methods of economical parameters

Given paper is devoted to some modern directions of the statistical analysis and forecasting of temporally series. On an example of two real temporary series, namely, rate of Ukrainian hrivna to US dollar and Russian ruble for the second half-year of 1997, the work of such methods of forecasting as a Brown method, method of exponential smoothing with using orthogonal Lagerr polynomials, adaptive Kalman filter is investigated. Also for these lines autoregression model of process is built, which is then checked on heteroscedasticity. The results of work of the foregoing methods, comparison of their opportunities are given in summary.