On-line sensitivity analysis of Markov chains
Continuous-time Markov processes on discrete state spaces (60J27) Queueing theory (aspects of probability theory) (60K25) Queues and service in operations research (90B22) Observability (93B07) Sensitivity (robustness) (93B35) Stochastic systems in control theory (general) (93E03) Estimation and detection in stochastic control theory (93E10)
The authors combine nominal and perturbed Markov chains in order to construct an augmented one. After eliminating the transient states, they transform this new chain into a reduced augmented chain with two properties: 1) Given the parameter change, the state probability perturbations may be expressed as simple linear combinations of the reduced augmented chain state probabilities. 2) The observability, i.e. the reduced augmented chain under certain conditions, can be efficiently estimated from the observations of the nominal Markov chain alone and this could be done on-line, i.e. while the nominal system is in operation. Applications to some queueing systems are included. The work is very interesting and useful.
- scientific article; zbMATH DE number 6520784
- Observable augmented systems for sensitivity analysis of Markov and semi-Markov processes
- Sensitivity analysis of stationary performance measures for Markov chains
- Sensitivity of the Stationary Distribution of a Markov Chain
- Sensitivity to perturbation of the stationary distribution: Some refinements
- Optimal inspection policies for a manufacturing station
- Perturbation analysis of discrete event systems: Concepts, algorithms, and applications
- Optimization via simulation: A review
- Massively parallel SIMD simulation of Markovian DEDS: Event and time synchronous methods
- Parallel sample path generation for discrete event systems and the traffic smoothing problem
- Distributed inversion in timed discrete event systems
- Variance reduction algorithms for parallel replicated simulation of uniformized Markov chains
- Sensitivity analysis of continuous time Bayesian network reliability models
- Smoothed perturbation analysis for queues with finite buffers
- A sample path analysis of M/GI/1 queues with workload restrictions
This page was built for publication: On-line sensitivity analysis of Markov chains
Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q3820510)