Šio darbo tikslas ištirti atsitiktinių procesų prognozavimo matematinius metodus ir nustačius nehomogeninio Puasono proceso algoritmą jį pritaikyti konkretiems atvejams. / In the first part of this work is theory, which focuses on random processes of prediction. In the second part of a non-homogeneous Poisson process explored prediction. In order to obtain the non-homogeneous Poisson process, we have a complete and sufficient statistics in three cases, the fourth complex cases such statistics didn’t exist. Taking advantage of these results were obtained by a non-homogeneous Poisson process uniformly minimum dispersion unbiased prediction in three of the cases examined, when the corresponding parameter is known or unknown. Along submitted analytical predictions expression.