Invariants of the Markov process by the transformation of variables

The guassian Markov process plays an important role in the identification and prediction of systems and the spectrum estimate of the time series. However, we are often confronted with the non-gaussian time series. In this paper, a method is discussed which transforms the data from non-gaussian to gaussian. It is shown that this procedure preserves the amount of Kullback information under the transformation. To verify the transformation statistically computer simulation was carried out.