Identifying an Unknown Process by Using Randomly Chosen Inputs.

Abstract

The authors investigated the problem of identifying the input-output relationship of an unknown deterministic process. In particular the authors investigated the ramifications of identifying solely on the basis of a finite number of input-output data pairs. Under this rather severe but realistic limitation the author have found identifications can be made by using randomly chosen inputs. Furthermore the authors have found that if two conditions are met, then the input-output relationship can always be constructed and therefore identified by using randomly chosen inputs. (Author)

Document Details

Document Type
Technical Report
Publication Date
Apr 10, 1972
Accession Number
AD0742062

Entities

People

  • Harold Stalford
  • Joseph Kullback

Organizations

  • United States Naval Research Laboratory

Tags

Communities of Interest

  • Materials and Manufacturing Processes

Readers

  • Adaptive Control and Estimation with Uncertainty in Dynamic Systems.
  • Regression Analysis.