A unified approach to Markov decision problems and performance sensitivity analysis (Q1974013)
From MaRDI portal
scientific article
| Language | Label | Description | Also known as |
|---|---|---|---|
| English | A unified approach to Markov decision problems and performance sensitivity analysis |
scientific article |
Statements
A unified approach to Markov decision problems and performance sensitivity analysis (English)
0 references
5 June 2001
0 references
The present note provides a unified formulation for the performance sensitivity analysis of Markov chains and Markov decision problems with both infinite horizon average-cost and discounted performance criteria. The approach is based on the concept of \(\alpha\)-potential (of a discounted Poisson equation), which can be estimated on the basis of a single-sample path of Markov chains.
0 references
discounted Lyapunov equations
0 references
perturbation analysis
0 references
sensitivity analysis of Markov chains
0 references
Markov decision problems
0 references
\(\alpha\)-potential
0 references
discounted Poisson equation
0 references
0 references
