This article is cited in 1 paper

Information Theory and Methods of Signal Processing

Error Bounds for the Method of Minimization of Empirical Risk

A. B. Tsybakov


Abstract: New nonasymptotic upper bounds are obtained for the probability of deviation of the minimum empirical risk from the minimum average risk in a class of problems that includes signal identification against an additive noise background under conditions of a priori uncertainty, as well as certain other pattern-recognition problems.

UDC: 621.391.1:519.27

Received: 25.12.1979

 English version: Problems of Information Transmission, 1981, 17:1, 33–42

Bibliographic databases:

• 
•