On-line learning of linear systems

S. E. Posner, S. R. Kulkarni

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Abstract

We consider the problem of learning/identification of a linear system by sampling its frequency response. A cumulative prediction error type criterion is used to describe the learnability of classes of (continuous- or discrete-time) linear systems which may have discontinuous frequency responses but of bounded variation. Upper and lower bounds are obtained for three input frequency sampling schemes: worst-case, random, and worst-case with small noise on the input frequency. Bounds are also obtained for the random sampling scheme with L1 noise or i.i.d. noise corrupting the frequency response samples.

Original languageEnglish (US)
Title of host publicationAmerican Control Conference
PublisherPubl by IEEE
Pages41-42
Number of pages2
ISBN (Print)0780308611, 9780780308619
DOIs
StatePublished - 1993
EventProceedings of the 1993 American Control Conference Part 3 (of 3) - San Francisco, CA, USA
Duration: Jun 2 1993Jun 4 1993

Publication series

NameAmerican Control Conference

Other

OtherProceedings of the 1993 American Control Conference Part 3 (of 3)
CitySan Francisco, CA, USA
Period6/2/936/4/93

All Science Journal Classification (ASJC) codes

  • Electrical and Electronic Engineering

Fingerprint

Dive into the research topics of 'On-line learning of linear systems'. Together they form a unique fingerprint.

Cite this