Learning MDPs from Features: Predict-Then-Optimize for Sequential Decision Problems by Reinforcement Learning

Abstract

In the predict-then-optimize framework, the objective is to train a predictive model, mapping from environment features to parameters of an optimization problem, which maximizes decision quality when the optimization is subsequently solved. Recent work on decision-focused learning shows that embedding the optimization problem in the training pipeline can improve decision quality and help generalize better to unseen tasks compared to relying on an intermediate loss function for evaluating prediction quality. We study the predict-then-optimize framework in the context of sequential decision problems (formulated as MDPs) that are solved via reinforcement learning. In particular, we are given environment features and a set of trajectories from training MDPs, which we use to train a predictive model that generalizes to unseen test MDPs without trajectories. Two significant computational challenges arise in applying decision-focused learning to MDPs: (i) large state and action spaces make it infeasible for existing techniques to differentiate through MDP problems, and (ii) the high-dimensional policy space, as parameterized by a neural network, makes differentiating through a policy expensive. We resolve the first challenge by sampling provably unbiased derivatives to approximate and differentiate through optimality conditions, and the second challenge by using a low-rank approximation to the high-dimensional sample-based derivatives. We implement both Bellman--based and policy gradient--based decision-focused learning on three different MDP problems with missing parameters, and show that decision-focused learning performs better in generalization to unseen tasks.

References in corpus (19)

• Model-Agnostic Meta-Learning for Fast Adaptation of Deep Networks
• Soft Actor-Critic: Off-Policy Maximum Entropy Deep Reinforcement Learning with a Stochastic Actor
• RL${}^2$: Fast Reinforcement Learning via Slow Reinforcement Learning
• Reinforcement Learning with Deep Energy-Based Policies
• Learning to reinforcement learn
• Optimization under Uncertainty in the Era of Big Data and Deep Learning: When Machine Learning Meets Mathematical Programming
• Deep Equilibrium Models
• Differentiable MPC for End-to-end Planning and Control
• MOReL : Model-Based Offline Reinforcement Learning
• Task-based End-to-end Model Learning in Stochastic Optimization
• VariBAD: A Very Good Method for Bayes-Adaptive Deep RL via Meta-Learning
• COMBO: Conservative Offline Model-Based Policy Optimization
• Interior Point Solving for LP-based prediction+optimisation
• Addressing the Loss-Metric Mismatch with Adaptive Loss Alignment
• Objective Mismatch in Model-based Reinforcement Learning
• POPCORN: Partially Observed Prediction COnstrained ReiNforcement Learning
• Collapsing Bandits and Their Application to Public Health Interventions
• Offline Meta-level Model-based Reinforcement Learning Approach for Cold-Start Recommendation
• Automatically Learning Compact Quality-aware Surrogates for Optimization Problems