RESTORATION OF DATA UNDER CONDITIONS OF INCOMPLETE SAMPLES

In studying adaptive or self‐organizing cybernetic systems problems arise of choosing an adequate model of the object and truth estimation of some initial data. In mathematical formulation many of these problems are reduced to the estimation of multidimensional system parameters under conditions of incomplete data. A problem is considered of estimation of multidimensional system parameters under conditions of a short sample with omissions. In the case when the data sample can be described by multidimensional normal law of a special type the method of the stated problem solution is suggested and grounded.