Influence appriory information about the initial object’s condition on accuracy of estimation of the model’s parameters

Abstract

The application of a Kalman’s filtering method is shown by development of algorithms of adaptive estimation in conditions of short sample. The technique of improvement of accuracy of estimation of the model’s parameters in the transitive period based on the account of the appriory information is described. The application of this technique for autoregression models and some dynamic lines in conditions of short sample is shown. The results of modeling of the offered algorithms are given.