• DocumentCode
    755358
  • Title

    Continuous-time recursive least-squares estimation, adaptive neural networks and systolic arrays

  • Author

    Dehaene, Jeroen ; Moonen, Marc ; Vandewalle, Joos

  • Volume
    42
  • Issue
    2
  • fYear
    1995
  • fDate
    2/1/1995 12:00:00 AM
  • Firstpage
    116
  • Lastpage
    119
  • Abstract
    We derive square-root covariance-type and information-type algorithms for continuous-time recursive least-squares estimation. The algorithms allow for easy manipulation and uniform parallelization. They are related to well-known neural adaptation laws and can be considered as continuous-time limits of systolic arrays
  • Keywords
    least squares approximations; neural nets; parallel algorithms; recursive estimation; systolic arrays; adaptive neural networks; continuous-time limits; continuous-time recursive estimation; covariance-type; information-type; neural adaptation laws; recursive least-squares estimation; square-root algorithms; systolic arrays; uniform parallelization; Adaptive systems; Covariance matrix; Iterative algorithms; Laboratories; Least squares approximation; Matrix decomposition; Neural networks; Recursive estimation; Solids; Systolic arrays;
  • fLanguage
    English
  • Journal_Title
    Circuits and Systems I: Fundamental Theory and Applications, IEEE Transactions on
  • Publisher
    ieee
  • ISSN
    1057-7122
  • Type

    jour

  • DOI
    10.1109/81.372852
  • Filename
    372852