Article contents
Programming problems and changes in the stable behavior of a class of Markov chains
Published online by Cambridge University Press: 14 July 2016
Abstract
This paper develops expressions for the derivatives with respect to a parameter μ of the stable probabilities of a class of Markov chains whose transition matrices are of the form Q + μW. These expressions lead to iterative schemes for calculation which in turn suggest gradient algorithms for finding locally optimal chains.
- Type
- Research Papers
- Information
- Copyright
- Copyright © Applied Probability Trust 1971
References
- 2
- Cited by