A lower bound on the performance of sequential prediction

Suleyman S. Kozat, Andrew C. Singer

Research output: Contribution to journalConference articlepeer-review


We consider the problem of sequential linear prediction of real-valued sequences under the square-error loss function. For this problem, a prediction algorithm has been demonstrated whose accumulated squared prediction error, for every bounded sequenee, is asymptotically as small as the best fixed linear predictor for that sequence, taken from the class of all linear predictors of a given order p. The redundancy, or excess prediction error above that of the best predictor for that sequence, is upper bounded by A2p ln(n)/n, where n is the data length and the sequence is assumed to be bounded by some A. In this paper, we show that this predictor is optimal in a min-max sense, by deriving a corresponding lower bound, such that no sequential predictor can ever do better than a redundancy of A2p ln(n)/n.

Original languageEnglish (US)
Pages (from-to)147
Number of pages1
JournalIEEE International Symposium on Information Theory - Proceedings
StatePublished - 2002
Event2002 IEEE International Symposium on Information Theory - Lausanne, Switzerland
Duration: Jun 30 2002Jul 5 2002

ASJC Scopus subject areas

  • Theoretical Computer Science
  • Information Systems
  • Modeling and Simulation
  • Applied Mathematics


Dive into the research topics of 'A lower bound on the performance of sequential prediction'. Together they form a unique fingerprint.

Cite this