A comparison of computational techniques of the key properties of Markov Chains

Date
2015-02-01
Authors
Hunter, JJ
Supervisor
Item type
Conference Contribution
Degree name
Journal Title
Journal ISSN
Volume Title
Publisher
Australia and New Zealand Industrial and Applied Mathematics (ANZIAM)
Abstract

The presenter has recently been exploring the accurate computation of the stationary distribution for finite Markov chains based upon the Grassman, Taksar and Heyman (GTH) algorithm ([1]) with further extensions of this procedure, based upon the ideas of Kohlas ([2]), for finding the mean first passage time matrix. The methods are numerically stable as they do not involve subtraction. In addition, a number of perturbation techniques, where the rows of the transition matrix are sequentially updated, are also considered for computing these quantities. These techniques, together with some standard techniques using matrix inverses and generalized inverses, are compared for accuracy, using some test problems from the literature. References: [1} Grassman W.K., Taksar M.I., and Heyman D.P., Regenerative analysis and steady state distributions for Markov chains, Oper. Res. 33, (1985), 1107-1116. [2] Kohlas J. Numerical computation of mean first passage times and absorption probabilities in Markov and semi-Markov models, Zeit fur Oper Res, 30, (1986), 197-207.

Description
Keywords
Source
Australia and New Zealand Industrial and Applied Mathematics Conference, (ANZIAM 15) held at Outrigger Surfers Paradise, Queensland, Australia, 2015-02-01 to 2015-02-05
DOI
Rights statement
NOTICE: this is the author’s version of a work that was accepted for publication. Changes resulting from the publishing process, such as peer review, editing, corrections, structural formatting, and other quality control mechanisms may not be reflected in this document. Changes may have been made to this work since it was submitted for publication. A definitive version was subsequently published in (see Citation). The original publication is available at (see Publisher's Version).