COMPARISON OF BAYESIAN APPROACH WITH CLASSICAL APPROACH FOR ESTIMATING THE PARAMETER OF MARKOV MODEL

  • Janardan Mahnata Department of Statistics, University of Chittagong, Chittagong, Bangladesh.
  • Soma Chowdhury Biswas Department of Statistics, University of Chittagong, Chittagong, Bangladesh.
Keywords: Credible Interval, Markov Model, Squared Error (SE), Tierney-Kadnae (T.K.)

Abstract

This paper has introduced Bayesian analysis and its application to estimate the parameter of the Markov model. To use Markov model, comparison between Bayesian approach and method of maximum likelihood have been done. Bayesian approach gives better result than classical approach. Jeffery’s non-informative prior and squared error loss function have been used in Bayesian inference. Tierney-Kadnae (T.K.) algorithm has been used to solve the Bayesian integral.

 

Downloads

Download data is not yet available.
Published
2016-12-16
Section
Articles