Paper: Modeling by shortest data description

  • Authors:
  • J. Rissanen

  • Affiliations:
  • IBM Research Laboratory, San Jose, California 95193, U.S.A.

  • Venue:
  • Automatica (Journal of IFAC)
  • Year:
  • 1978

Quantified Score

Hi-index 22.17

Visualization

Abstract

The number of digits it takes to write down an observed sequence x"1, ..., x"N of a time series depends on the model with its parameters that one assumes to have generated the observed data. Accordingly, by finding the model which minimizes the description length one obtains estimates of both the integer-valued structure parameters and the real-valued system parameters.