MATRIX PREDICTIVE MODEL AND ITS LEARNING IN TASKS OF ENVIRONMENTAL MONITORING
Abstract
An approach to the synthesis of matrix predictive models of two-dimensional fields of observation is offeredandintroduced the procedure of estimating the parameters of these models which are generalizations of adaptive algorithms identify the matrix case,for the solution of problems of environmental monitoring, that have a problem related to the fact that the raw data are"noised" by intense interference and controlled by themselves sufficient non-stationary sequences. These procedures are characterized by simple numerical implementation.