More Like this

1. Solving Markov decision processes for network-level post-hazard recovery via simulation optimization and rollout

   Sarkale, Y.; Nozhati, S.; Chong, E. K.; Ellingwood, B. R.; Mahmoud, H. (August 2019, 2018 IEEE 14th International Conference on Automation Science and Engineering (CASE))

   Computation of optimal recovery decisions for community resilience assurance post-hazard is a combinatorial decision-making problem under uncertainty. It involves solving a large-scale optimization problem, which is significantly aggravated by the introduction of uncertainty. In this paper, we draw upon established tools from multiple research communities to provide an effective solution to this challenging problem. We provide a stochastic model of damage to the water network (WN) within a testbed community following a severe earthquake and compute near-optimal recovery actions for restoration of the water network. We formulate this stochastic decision-making problem as a Markov Decision Process (MDP), and solve it using a popular class of heuristic algorithms known as rollout. A simulation-based representation of MDPs is utilized in conjunction with rollout and the Optimal Computing Budget Allocation (OCBA) algorithm to address the resulting stochastic simulation optimization problem. Our method employs non-myopic planning with efficient use of simulation budget. We show, through simulation results, that rollout fused with OCBA performs competitively with respect to rollout with total equal allocation (TEA) at a meagre simulation budget of 5-10% of rollout with TEA, which is a crucial step towards addressing large-scale community recovery problems following natural disasters. 

   more » « less
2. Proximal Reinforcement Learning: Efficient Off-Policy Evaluation in Partially Observed Markov Decision Processes

   https://doi.org/10.1287/opre.2021.0781  

   Bennett, Andrew; Kallus, Nathan (September 2023, Operations Research)

   In applications of offline reinforcement learning to observational data, such as in healthcare or education, a general concern is that observed actions might be affected by unobserved factors, inducing confounding and biasing estimates derived assuming a perfect Markov decision process (MDP) model. In “Proximal Reinforcement Learning: Efficient Off-Policy Evaluation in Partially Observed Markov Decision Processes,” A. Bennett and N. Kallus tackle this by considering off-policy evaluation in a partially observed MDP (POMDP). Specifically, they consider estimating the value of a given target policy in an unknown POMDP, given observations of trajectories generated by a different and unknown policy, which may depend on the unobserved states. They consider both when the target policy value can be identified the observed data and, given identification, how best to estimate it. Both these problems are addressed by extending the framework of proximal causal inference to POMDP settings, using sequences of so-called bridge functions. This results in a novel framework for off-policy evaluation in POMDPs that they term proximal reinforcement learning, which they validate in various empirical settings. 

   more » « less
3. Planning with Abstract Markov Decision Processes

   Gopalan, Nakul; desJardins, Marie; Littman, Michael L.; MacGlashan, J.; Squire, S.; Tellex, Stefanie; Winder, John; Wong, Lawson L. (July 2017, 27th International Conference on Automated Planning and Scheduling)

   Robots acting in human-scale environments must plan under uncertainty in large state–action spaces and face constantly changing reward functions as requirements and goals change. Planning under uncertainty in large state–action spaces requires hierarchical abstraction for efficient computation. We introduce a new hierarchical planning framework called Abstract Markov Decision Processes (AMDPs) that can plan in a fraction of the time needed for complex decision making in ordinary MDPs. AMDPs provide abstract states, actions, and transition dynamics in multiple layers above a base-level “flat” MDP. AMDPs decompose problems into a series of subtasks with both local reward and local transition functions used to create policies for subtasks. The resulting hierarchical planning method is independently optimal at each level of abstraction, and is recursively optimal when the local reward and transition functions are correct. We present empirical results showing significantly improved planning speed, while maintaining solution quality, in the Taxi domain and in a mobile-manipulation robotics problem. Furthermore, our approach allows specification of a decision-making model for a mobile-manipulation problem on a Turtlebot, spanning from low-level control actions operating on continuous variables all the way up through high-level object manipulation tasks. 

   more » « less
4. Planning with Abstract Markov Decision Processes

   Gopalan, Nakul; desJardins, Marie; Littman, Michael L.; MacGlashan, James; Squire, Shawn; Tellex, Stefanie; Winder, John; Wong, Lawson L.S. (January 2017, ICAPS)

   Robots acting in human-scale environments must plan under uncertainty in large state–action spaces and face constantly changing reward functions as requirements and goals change. Planning under uncertainty in large state–action spaces requires hierarchical abstraction for efficient computation. We introduce a new hierarchical planning framework called Abstract Markov Decision Processes (AMDPs) that can plan in a fraction of the time needed for complex decision making in ordinary MDPs. AMDPs provide abstract states, actions, and transition dynamics in multiple layers above a base-level “flat” MDP. AMDPs decompose problems into a series of subtasks with both local reward and local transition functions used to create policies for subtasks. The resulting hierarchical planning method is independently optimal at each level of abstraction, and is recursively optimal when the local reward and transition functions are correct. We present empirical results showing significantly improved planning speed, while maintaining solution quality, in the Taxi domain and in a mobile-manipulation robotics problem. Furthermore, our approach allows specification of a decision-making model for a mobile-manipulation problem on a Turtlebot, spanning from low-level control actions operating on continuous variables all the way up through high-level object manipulation tasks. 

   more » « less
5. Robust Average-Reward Markov Decision Processes

   https://doi.org/10.1609/aaai.v37i12.26775  

   Wang, Yue; Velasquez, Alvaro; Atia, George; Prater-Bennette, Ashley; Zou, Shaofeng (June 2023, Proceedings of the AAAI Conference on Artificial Intelligence)

   In robust Markov decision processes (MDPs), the uncertainty in the transition kernel is addressed by finding a policy that optimizes the worst-case performance over an uncertainty set of MDPs. While much of the literature has focused on discounted MDPs, robust average-reward MDPs remain largely unexplored. In this paper, we focus on robust average-reward MDPs, where the goal is to find a policy that optimizes the worst-case average reward over an uncertainty set. We first take an approach that approximates average-reward MDPs using discounted MDPs. We prove that the robust discounted value function converges to the robust average-reward as the discount factor goes to 1, and moreover when it is large, any optimal policy of the robust discounted MDP is also an optimal policy of the robust average-reward. We further design a robust dynamic programming approach, and theoretically characterize its convergence to the optimum. Then, we investigate robust average-reward MDPs directly without using discounted MDPs as an intermediate step. We derive the robust Bellman equation for robust average-reward MDPs, prove that the optimal policy can be derived from its solution, and further design a robust relative value iteration algorithm that provably finds its solution, or equivalently, the optimal robust policy. 

   more » « less