Dynamic Programming for Epistemic Uncertainty in Markov Decision Processes
Preprint
Abstract
In this paper, we propose a general theory of ambiguity-averse MDPs, which treats the uncertain transition probabilities as random variables and evaluates a policy via a risk measure applied to its random return. This ambiguity-averse MDP framework unifies several models of MDPs with epistemic uncertainty for specific choices of risk measures. We extend the concepts of value functions and Bellman operators to our setting. Based on these objects, we establish the consequences of dynamic programming principles in this framework (existence of stationary policies, value and policy iteration algorithms), and we completely characterize law-invariant risk measures compatible with dynamic programming. Our work draws connections among several variants of MDP models and fully delineates what is possible under the dynamic programming paradigm and which risk measures require leaving it.Keywords
All Related Versions
This publication has 16 references indexed in Scilit:
- Robust Markov Decision Processes: Beyond RectangularityMathematics of Operations Research, 2023
- “Dice”-sion–Making Under Uncertainty: When Can a Random Decision Reduce Risk?Management Science, 2019
- Computation of weighted sums of rewards for concurrent MDPsMathematical Methods of Operations Research, 2018
- Tight Approximations of Dynamic Risk MeasuresMathematics of Operations Research, 2015
- Time consistency of dynamic risk measuresOperations Research Letters, 2012
- Distributionally Robust Markov Decision ProcessesMathematics of Operations Research, 2012
- Markov Decision Processes with Average-Value-at-Risk criteriaMathematical Methods of Operations Research, 2011
- Representation results for law invariant time consistent functionsMathematics and Financial Economics, 2009
- COHERENT ACCEPTABILITY MEASURES IN MULTIPERIOD MODELSMathematical Finance, 2005
- Stochastic finance. an introduction in discrete timeThe Mathematical Intelligencer, 2004