Variational characterizations in Markov decision processes (Q1077334): Difference between revisions

Variational characterizations and bounds for the solutions of systems of functional equations in discounted and undiscounted semi-Markov decision processes are obtained. Such upper and lower bounds can be used to measure the deviation of the current solution from optimality. The variational characterizations suggest numerical algorithms.

0 references

zbMATH Keywords

Variational characterizations

0 references

discounted and undiscounted semi-Markov decision processes

0 references

0 references

0 references

Contraction Mappings in the Theory Underlying Dynamic Programming

0 references

Markov Renewal Programs with Small Interest Rates

0 references

Multichain Markov Renewal Programs

0 references

Q3206684

0 references

Successive Approximation Methods for Solving Nested Functional Equations in Markov Decision Problems

0 references

Technical Note—Elimination of Suboptimal Actions in Markov Decision Problems

0 references

Technical Note—Bounds on the Gain of a Markov Decision Process

0 references

Note—A Test for Nonoptimal Actions in Undiscounted Finite Markov Decision Chains

0 references

Tests for Suboptimal Actions in Discounted Markov Programming

0 references

Linear Programming and Markov Decision Chains

0 references

Q3266141

0 references

Markov-Renewal Programming. I: Formulation, Finite Return Models

0 references

Potentials for denumerable Markov chains

0 references

A modified dynamic programming method for Markovian decision problems

0 references

Letter to the Editor—A Test for Suboptimal Actions in Markovian Decision Problems

0 references

Discrete Dynamic Programming with a Small Interest Rate

0 references

Q5821083

0 references

On Finding the Maximal Gain for Markov Decision Processes

0 references

Some Bounds for Discounted Sequential Decision Processes

0 references

Action Elimination Procedures for Modified Policy Iteration Algorithms

0 references

On the solvability of Bellman's functional equation for a Markovian decision process

0 references

Perturbation theory and finite Markov chains

0 references

Iterative solution of the functional equations of undiscounted Markov renewal programming

0 references

On the solvability of Bellman's functional equations for Markov renewal programming

0 references

The Functional Equations of Undiscounted Markov Renewal Programming

0 references

Geometric convergence of value-iteration in multichain Markov decision problems

0 references

Discrete Dynamic Programming with Sensitive Discount Optimality Criteria

0 references

Q3908795

0 references

A UNIFIED APPROACH TO ALGORITHMS WITH A SUBOPTIMALITY TEST IN DISCOUNTED SEMI-MARKOV DECISION PROCESSES

0 references

Identifiers

zbMATH Open document ID

0594.90088

0 references

DOI

10.1016/0022-247X(86)90229-5

0 references

Mathematics Subject Classification ID

0 references

0 references

0 references

0 references

Sitelinks

Mathematics(1 entry)

mardi Publication:1077334

@@ Property / author @@
+Awi Federgruen
@@ Property / author: Awi Federgruen / rank @@
+Normal rank
@@ Property / reviewed by @@
+Howard J. Weiner
@@ Property / reviewed by: Howard J. Weiner / rank @@
+Normal rank
@@ Property / MaRDI profile type @@
+Publication
@@ Property / MaRDI profile type: Publication / rank @@
+Normal rank
@@ Property / cites work @@
+Contraction Mappings in the Theory Underlying Dynamic Programming
+Normal rank
@@ Property / cites work @@
+Markov Renewal Programs with Small Interest Rates
@@ Property / cites work: Markov Renewal Programs with Small Interest Rates / rank @@
+Normal rank
@@ Property / cites work @@
+Multichain Markov Renewal Programs
@@ Property / cites work: Multichain Markov Renewal Programs / rank @@
+Normal rank
@@ Property / cites work @@
+Q3206684
@@ Property / cites work: Q3206684 / rank @@
+Normal rank
@@ Property / cites work @@
+Successive Approximation Methods for Solving Nested Functional Equations in Markov Decision Problems
+Normal rank
@@ Property / cites work @@
+Technical Note—Elimination of Suboptimal Actions in Markov Decision Problems
+Normal rank
@@ Property / cites work @@
+Technical Note—Bounds on the Gain of a Markov Decision Process
+Normal rank
@@ Property / cites work @@
+Note—A Test for Nonoptimal Actions in Undiscounted Finite Markov Decision Chains
+Normal rank
@@ Property / cites work @@
+Tests for Suboptimal Actions in Discounted Markov Programming
+Normal rank
@@ Property / cites work @@
+Linear Programming and Markov Decision Chains
@@ Property / cites work: Linear Programming and Markov Decision Chains / rank @@
+Normal rank
@@ Property / cites work @@
+Q3266141
@@ Property / cites work: Q3266141 / rank @@
+Normal rank
@@ Property / cites work @@
+Markov-Renewal Programming. I: Formulation, Finite Return Models
+Normal rank
@@ Property / cites work @@
+Potentials for denumerable Markov chains
@@ Property / cites work: Potentials for denumerable Markov chains / rank @@
+Normal rank
@@ Property / cites work @@
+A modified dynamic programming method for Markovian decision problems
+Normal rank
@@ Property / cites work @@
+Letter to the Editor—A Test for Suboptimal Actions in Markovian Decision Problems
+Normal rank
@@ Property / cites work @@
+Discrete Dynamic Programming with a Small Interest Rate
+Normal rank
@@ Property / cites work @@
+Q5821083
@@ Property / cites work: Q5821083 / rank @@
+Normal rank
@@ Property / cites work @@
+On Finding the Maximal Gain for Markov Decision Processes
+Normal rank
@@ Property / cites work @@
+Some Bounds for Discounted Sequential Decision Processes
+Normal rank
@@ Property / cites work @@
+Action Elimination Procedures for Modified Policy Iteration Algorithms
+Normal rank
@@ Property / cites work @@
+On the solvability of Bellman's functional equation for a Markovian decision process
+Normal rank
@@ Property / cites work @@
+Perturbation theory and finite Markov chains
@@ Property / cites work: Perturbation theory and finite Markov chains / rank @@
+Normal rank
@@ Property / cites work @@
+Iterative solution of the functional equations of undiscounted Markov renewal programming
+Normal rank
@@ Property / cites work @@
+On the solvability of Bellman's functional equations for Markov renewal programming
+Normal rank
@@ Property / cites work @@
+The Functional Equations of Undiscounted Markov Renewal Programming
+Normal rank
@@ Property / cites work @@
+Geometric convergence of value-iteration in multichain Markov decision problems
+Normal rank
@@ Property / cites work @@
+Discrete Dynamic Programming with Sensitive Discount Optimality Criteria
+Normal rank
@@ Property / cites work @@
+Q3908795
@@ Property / cites work: Q3908795 / rank @@
+Normal rank
@@ Property / cites work @@
+A UNIFIED APPROACH TO ALGORITHMS WITH A SUBOPTIMALITY TEST IN DISCOUNTED SEMI-MARKOV DECISION PROCESSES
+Normal rank