Optimal choice of the moments of observation in a problem of distinguishing stochastic processes
- 15 Downloads
The problem is considered of distinguishing two Wiener processes with known diffusion coefficients on the basis of a finite number of inexact observations on a given time interval. In connection with the optimal choice of the moments of observation, the asymptotics of the maximum of the entropy distance of the arising pairs of finite-dimensional Gaussian distributions are found; the question of the optimal choice of the moments of observations is discussed, and the behavior of the entropy distance is studied for a fixed number of observations when the accuracy is increased.
KeywordsEntropy Gaussian Distribution Diffusion Coefficient Stochastic Process Finite Number
Unable to display preview. Download preview PDF.
- 1.Yu. A. Rozanova, “Infinite-dimensional Gaussian distributions,” Tr. Mat. Inst. im. V. A. Steklova Akad. Nauk SSSR,108 (1968).Google Scholar
- 2.N. N. Chentsov, Statistical Solution Rules and Optimal Deductions [in Russian], Nauka, Moscow (1972).Google Scholar
- 3.U. Grenander and G. Szegö, Töplitz Forms and Their Applications [Russian translation], Moscow (1981).Google Scholar