Yuriy A.
Belov Andrey L.
Zavorotnyi Veda S.
Kasyanyuk ABSTRACT Two methods of linear functional value estimation by noised values of other linear functionals with unknown kernels have been developed using multi-criteria optimization idea. The behavior of optimization criteria under variation of pareto-optimum parameter are analyzed and the results are presented. Estimation quality and speed of these two methods were compared. Estimation quality dependence on quantity of test signals was shown. Some problems of pareto-optimum parameter choice were solved. Model problem program-solved results using the developed methods were submitted.
<< Previous article Next article >> |
||||||||||||||||||||||