| Abstract: | While complex systems are being designed, a number of statistical experiments are carried out. The present paper is devoted to the development of methods of correlated processes which increase the efficiency of statistical tests by designing dependent experiments and simultaneous processing of the obtained data. The deficiency of the usual approaches in this field consists in their specialization in complex systems modelled by functionals of the mathematical expectaition only. We consider here the technique of minimal contrast estimation which allows us to extend the methods of correlated processes to a large family of functionals. Translated fromStatisticheskie Metody Otsenivaniya i Proverki Gipotez, pp. 150–163, Perm, 1990. |
