|
ON GUARANTEED ESTIMATION OF PARAMETERS OF RANDOM PROCESSES BY THE WEIGHTED LEAST SQUARES METHOD
Nataliya Meder and Sergei Vorobejchikov
Department of Applied Mathematics and Cybernetics,
Tomsk University, 36 Lenin str., 634050 Tomsk, Russia
e-mail: meder_n@mail.ru
The problem of parameters estimation of an autoregressive process is considered. The method of guaranteed estimation is based on the least squares method with special weights and uses a special stopping rule. The properties of the procedures are studied for the case of known and unknown variance of the noise.
Keywords: Time series analysis, autoregressive models, sequential estimation, least-square method, mean square error, numerical simulation
Session slot T-Fr-A01: Identification of Stochastic Systems/Area code 3a : Modelling, Identification and Signal Processing
 | 