• DocumentCode
    2835320
  • Title

    A new method for estimating Markov random fields

  • Author

    Almeida, Murilo ; Gidas, Basilis

  • Author_Institution
    Div. of Appl. Math., Brown Univ., Providence, RI, USA
  • fYear
    1989
  • fDate
    22-24 Nov 1989
  • Firstpage
    340
  • Lastpage
    341
  • Abstract
    A method is introduced for estimating the parameters of Markov random fields from complete or incomplete data. The method is computationally more efficient than its best competitor, the method of maximum pseudo-likelihood. For incomplete data, the method leads to an algorithm which is similar to, but simpler than, the EM algorithm
  • Keywords
    Markov processes; data analysis; parameter estimation; Markov random fields; incomplete data; parameter estimation; random field estimation; variational method; Computer network reliability; Computer networks; Computer vision; Degradation; Markov random fields; Mathematics; Neural networks; Parameter estimation; Speech recognition; Virtual manufacturing;
  • fLanguage
    English
  • Publisher
    ieee
  • Conference_Titel
    TENCON '89. Fourth IEEE Region 10 International Conference
  • Conference_Location
    Bombay
  • Type

    conf

  • DOI
    10.1109/TENCON.1989.176881
  • Filename
    176881
    • Link To Document
    https://search.isc.ac/dl/search/defaultta.aspx?DTC=49&DC=2835320