The Computation of Stationary Distributions on Markov Chains through Perturbations.
Abstract
An algorithmic procedure, for the determination of the stationary distribution of a finite, m-state, irreducible Markov chain, that does not require the use of methods for solving systems of linear equations, is presented. The technique is based upon a succession of m, rank one, perturbations of the trivial doubly stochastic matrix whose known steady state vector is updated at each stage to yield the required stationary probability vector.
Document Details
- Document Type
- Technical Report
- Publication Date
- Mar 01, 1988
- Accession Number
- ADA193191
Entities
People
- Jeffrey J. Hunter
Organizations
- University of North Carolina at Chapel Hill