The method of non-linear lag correlation rejection from the non-stationary time series is constructed. This method is based on the analysis of support of the combined sample distribution function density of lag vector components. The method of correlation analysis between two time series is also formulated. The concrete forecast example of non-stationary time series is given for the case of definite level of stationary state of sample distribution function.

Mathematical modelling in actual problems of science and technics

Bitter V.V.

Kirillov D.S.

Korob O.V.

Orlov Yurii Nikolaevich,  ,  orcid.org/0000-0002-1356-5137,  KIAM RAS

Russkov A.A.