In this paper we propose the prognosis of the unemployment rate in the European Union through the Box-Jenkins method and the TRAMO/SEATS method as well as the detection of the method which proves to provide the best results. The monthly unemployment rate in the European Union is affected by seasonal variations of deterministic and stochastic nature. The prognosis through the Box-Jenkins nature supposes the separate consideration of seasonal variations, according to their specific nature. The stochastic seasonal variations are modelled and prognosticated simultaneously with the other components of the time series, based on the generating stochastic process. The prognosis of the monthly unemployment rate in the European Union through the TRAMO/SEATS methods is done by aggregating the individual prognoses of the components of the time series, obtained according to the stochastic processes models that generate them.

seasonal variations, stochastic process, moving average, prognosis, performance indicators of the prognosis