Using adaptive hybrid models in forecasting foreign outflows from Russia

The quality of migration forecasts from Russia to foreign countries, constructed using exponential smoothing methods, autoregressive integrated moving average (ARIMA) and «naive» extrapolation is compared. For retrospective testing of forecasts, data on outflows from Russia to 53 countries of the world and a group of states «other countries» were used. The obtained results indicate that the use of hybrid models synthesizing the above methods can improve the quality of the forecast.