Optimization of algorithms for estimating parameters of stochastic systems in conditions of indeterminacy (Q1072992)

The problem of estimating the parameters of a vector regression model with correlated measurement errors is considered. A general estimation method is proposed, whereby, given imperfect information about the correlation structure of the measurement errors, an adaptive estimate can be obtained for the unknown parameters, which is asymptotically not less accurate than the best linear unbiased estimate. The statistical properties of the estimate are examined, and the results of numerical experiments are quoted.