Adaptive Systems

On the empiric risk minimization method in identification problems

A. B. Tsybakov

Moscow

Abstract: The paper is concerned with asymptotic properties of the method for minimization of empirical risk in identification problems. Consistency, asymptotic normalcy, and convergence conditions are established for moments of parameter estimates of some nonlinear models.

UDC: 62-501.7


Received: 23.06.1980

 English version: Automation and Remote Control, 1981, 42:9, 1196–1203

Bibliographic databases:

• 
•