More Like this

1. Future-Dependent Value-Based Off-Policy Evaluation in POMDPs

   Uehara, Masatoshi; Kiyohara, Haruka; Bennett, Andrew; Chernozhukov, Victor; Jiang, Nan; Kallus, Nathan; Shi, Chengchun; Sun, Wen (December 2023, Neural Information Processing Systems (NeurIPS 2023))

   We study off-policy evaluation (OPE) for partially observable MDPs (POMDPs) with general function approximation. Existing methods such as sequential im- portance sampling estimators suffer from the curse of horizon in POMDPs. To circumvent this problem, we develop a novel model-free OPE method by introduc- ing future-dependent value functions that take future proxies as inputs and perform a similar role to that of classical value functions in fully-observable MDPs. We derive a new off-policy Bellman equation for future-dependent value functions as conditional moment equations that use history proxies as instrumental variables. We further propose a minimax learning method to learn future-dependent value functions using the new Bellman equation. We obtain the PAC result, which implies our OPE estimator is close to the true policy value under Bellman completeness, as long as futures and histories contain sufficient information about latent states. 

   more » « less
2. Towards optimal off-policy evaluation for reinforcement learning with marginalized importance sampling.

   Xie, T.; Ma, Y.; Wang, Y. X. (January 2019, Advances in neural information processing systems)

   Motivated by the many real-world applications of reinforcement learning (RL) that require safe-policy iterations, we consider the problem of off-policy evaluation (OPE) — the problem of evaluating a new policy using the historical data ob- tained by different behavior policies — under the model of nonstationary episodic Markov Decision Processes (MDP) with a long horizon and a large action space. Existing importance sampling (IS) methods often suffer from large variance that depends exponentially on the RL horizon H. To solve this problem, we consider a marginalized importance sampling (MIS) estimator that recursively estimates the state marginal distribution for the target policy at every step. MIS achieves a mean-squared error of [ ] where μ and π are the logging and target policies, dμt (st) and dπt (st) are the marginal distribution of the state at tth step, H is the horizon, n is the sample size and V π is the value function of the MDP under π. The result matches the t+1 Cramer-Rao lower bound in Jiang and Li [2016] up to a multiplicative factor of H. To the best of our knowledge, this is the first OPE estimation error bound with a polynomial dependence on H . Besides theory, we show empirical superiority of our method in time-varying, partially observable, and long-horizon RL environments. 

   more » « less
3. Sampling Networks and Aggregate Simulation for Online POMDP Planning

   Cui, H; Khardon, R. (December 2019, Advances in neural information processing systems)

   The paper introduces a new algorithm for planning in partially observable Markov decision processes (POMDP) based on the idea of aggregate simulation. The algorithm uses product distributions to approximate the belief state and shows how to build a representation graph of an approximate action-value function over belief space. The graph captures the result of simulating the model in aggregate under independence assumptions, giving a symbolic representation of the value function. The algorithm supports large observation spaces using sampling networks, a representation of the process of sampling values of observations, which is integrated into the graph representation. Following previous work in MDPs this approach enables action selection in POMDPs through gradient optimization over the graph representation. This approach complements recent algorithms for POMDPs which are based on particle representations of belief states and an explicit search for action selection. Our approach enables scaling to large factored action spaces in addition to large state spaces and observation spaces. An experimental evaluation demonstrates that the algorithm provides excellent performance relative to state of the art in large POMDP problems. 

   more » « less
4. Provably Efficient Reinforcement Learning in Partially Observable Dynamical Systems

   Masatoshi Uehara, Ayush Sekhari (January 2022, Advances in neural information processing systems)

   We study Reinforcement Learning for partially observable systems using function approximation. We propose a new PO-bilinear framework, that is general enough to include models such as undercomplete tabular Partially Observable Markov Decision Processes (POMDPs), Linear Quadratic Gaussian (LQG), Predictive State Representations (PSRs), as well as a newly introduced model Hilbert Space Embeddings of POMDPs. Under this framework, we propose an actor-critic style algorithm that is capable to performing agnostic policy learning. Given a policy class that consists of memory based policies (i.e., policy that looks at a fixed-length window of recent observations), and a value function class that consists of functions taking both memory and future observations as inputs, our algorithm learns to compete against the best memory-based policy among the policy class. For certain examples such as undercomplete POMDPs and LQGs, by leveraging their special properties, our algorithm is even capable of competing against the globally optimal policy without paying an exponential dependence on the horizon. 

   more » « less
5. Efficiently Breaking the Curse of Horizon in Off-Policy Evaluation with Double Reinforcement Learning

   https://doi.org/10.1287/opre.2021.2249  

   Kallus, Nathan; Uehara, Masatoshi (January 2022, Operations Research)

   Off-policy evaluation (OPE) in reinforcement learning is notoriously difficult in long- and infinite-horizon settings due to diminishing overlap between behavior and target policies. In this paper, we study the role of Markovian and time-invariant structure in efficient OPE. We first derive the efficiency bounds and efficient influence functions for OPE when one assumes each of these structures. This precisely characterizes the curse of horizon: in time-variant processes, OPE is only feasible in the near-on-policy setting, where behavior and target policies are sufficiently similar. But, in time-invariant Markov decision processes, our bounds show that truly off-policy evaluation is feasible, even with only just one dependent trajectory, and provide the limits of how well we could hope to do. We develop a new estimator based on double reinforcement learning (DRL) that leverages this structure for OPE. Our DRL estimator simultaneously uses estimated stationary density ratios and q-functions and remains efficient when both are estimated at slow, nonparametric rates and remains consistent when either is estimated consistently. We investigate these properties and the performance benefits of leveraging the problem structure for more efficient OPE. 

   more » « less