New algorithms solve robust MDPs efficiently, significantly faster than existing methods.
problem Computing robust MDP solutions with uncertainty in transition probabilities is computationally expensive.
method Partial policy iteration and fast robust Bellman operator computation methods.
result The proposed methods are many orders of magnitude faster than state-of-the-art approaches.
Improved RL algorithm with linear MDPs for offline learning with partial data coverage.
problem Efficient offline RL with linear MDPs under partial data coverage.
method Primal-dual algorithm with O ( ε − 2 ) O(ε^{-2}) O ( ε − 2 ) sample complexity. result First computationally efficient algorithm with O ( ε − 2 ) O(ε^{-2}) O ( ε − 2 ) sample complexity for offline RL with linear MDPs under partial data coverage. Paper analyzes history-based RL methods for MDPs, introduces a theoretical framework and practical algorithm.
problem Improving RL performance in MDPs using history-based features.
method Theoretical framework for history-based RL, practical algorithm design.
result Practical RL algorithm shows effectiveness on continuous control tasks.
We characterize value functions in partially observable MDPs as semi-algebraic sets.
problem Understanding feasible value functions in partially observable Markov decision processes.
method Characterization of feasible value functions as semi-algebraic sets defined by polynomial inequalities.
result The feasible set of value functions in POMDPs is a semi-algebraic set, not a polytope as in MDPs.
Algorithm improves reinforcement learning in MDPs with partial order policies.
problem Improving reinforcement learning in MDPs with partial order policies.
method Epoch-based reinforcement learning algorithm leveraging a partial order over policy class.
result Achieves an O ( w log ( ∣ Θ ∣ ) T ) O(\sqrt{w \log(|Θ|) T}) O ( w log ( ∣Θ∣ ) T ) regret bound, independent of state and action space sizes. CPPO learns policies from partial offline data in MDPs with structural assumptions.
problem Offline Reinforcement Learning with partial coverage assumption.
method Constrained Pessimistic Policy Optimization (CPPO) using a function class and model class constraint.
result CPPO achieves PAC guarantee with partial coverage, learning competitive policies.
MetaCURL tackles non-stationary MDPs with optimal dynamic regret.
problem Online learning in non-stationary Markov decision processes.
method MetaCURL uses a meta-algorithm with multiple black-box algorithms and a sleeping expert framework.
result Achieves optimal dynamic regret without prior knowledge of MDP changes.
Transformer learns to infer partial MDPs for efficient in-context adaptation and exploration.
problem Efficiently adapt and explore in-context without gradient-based updates.
method Uses a transformer to learn inference from training tasks, considering hypothesis space of partial models.
result Adaptation speed and exploration-exploitation balance approach those of an exact posterior sampling oracle.
RL approach for target tracking with unknown dynamics and sensor control.
problem Tracking an unknown target with sensor control.
method Track-MDP formulation for RL, compared with POMDP.
result Optimal RL policy tracks all target paths with certainty.
PRL improves off-policy evaluation in partially observed MDPs.
problem Confounding and bias in offline reinforcement learning with unobserved state factors.
method Extends proximal causal inference to POMDPs, identifying and estimating target policy value.
result Semiparametrically efficient estimators for PRL in partially observed MDPs.
We solve POMDPs by approximating them as finite-state MDPs.
problem Computational challenges in learning optimal policies for POMDPs.
method Transform POMDP into a Superstate MDP, apply TD-learning and policy optimization.
result Finite-time bounds on TD-learning error for non-Markovian dynamics.
New method approximates POMDPs with PB-MDPs, providing error bounds and practical algorithms.
problem Difficulty in solving POMDPs with continuous or hybrid state and observation spaces.
method Bounding particle filtering error and adapting MDP algorithms to POMDPs.
result General theory and practical algorithms for POMDPs with no direct dependence on state and observation space sizes.
New framework for reinforcement learning with sporadic state observations.
problem Partial observability in reinforcement learning.
method Action-Triggered Sporadically Traceable Markov Decision Processes (ATST-MDPs).
result Optimistic algorithm achieving regret bound for episodic learning.
New algorithm learns optimal policies in strategic MDPs with private types.
problem Optimal policy learning in strategic MDPs with private types and information asymmetry.
method PLAN algorithm using instrumental variable regression and pessimism principle.
result PLAN achieves near-optimal policy with 1 / K 1 / \sqrt{K} 1/ K optimality. Unified framework for efficient RL in MDP, POMDP, and PSR.
problem Sample efficient reinforcement learning under various decision-making frameworks.
method Proposed GEC complexity measure and a posterior sampling algorithm.
result Sample efficient learning achieved with sublinear regret bound in terms of GEC.
Study shows offline RL with partial coverage and weak function classes is possible.
problem Learning optimal policies from logged data with function approximation.
method Marginalized Importance Sampling (MIS) with additional covering distribution.
result Finite-sample guarantees for sample-efficient offline RL for general MDPs.
Humans achieve efficient learning by relying on prior knowledge about the structure of naturally occurring tasks. There is considerable interest in designing reinforcement learning (RL) algorithms with similar properties. This includes proposals to learn the learning algorithm itself, an idea also known as meta learnin…
This work tackles representation learning for RL in low-rank MDPs, improving sample efficiency.
problem How to learn a compact representation for RL in low-rank MDPs efficiently.
method Proposes REP-UCB for online RL and develops an algorithm for offline RL under partial coverage.
result Significantly improved sample complexity for online RL and competitive performance for offline RL.
CycleFQI tackles offline reinforcement learning for cyclic MDPs, mitigating state distribution mismatch.
problem Offline reinforcement learning for cyclic MDPs with heterogeneous dynamics and discount factors.
method CycleFQI decomposes the cyclic process into stage-wise sub-problems, using vector of stage-specific Q-functions.
result CycleFQI mitigates the curse of dimensionality and provides finite-sample suboptimality error bounds.
Reinforcement Learning (RL) is a learning paradigm concerned with learning to control a system so as to maximize an objective over the long term. This approach to learning has received immense interest in recent times and success manifests itself in the form of human-level performance on games like \textit{Go}. While R…
OMLE combines optimism and MLE for efficient sequential decision making.
problem Efficiently solving sequential decision making problems, especially in partially observable settings.
method Combines optimism for exploration and maximum likelihood estimation for model learning.
result OMLE learns near-optimal policies for a wide range of sequential decision making problems.
New coverage conditions improve sample efficiency in online reinforcement learning.
problem Improving sample efficiency in online reinforcement learning with function approximation.
method Identifying and studying new coverage conditions for online reinforcement learning.
result Improved regret bounds achieved with new coverage conditions.
Study on future-dependent value functions for off-policy evaluation in complex environments.
problem Exponential dependence on horizon in off-policy evaluation for complex observations.
method Developed novel coverage assumptions for POMDPs to achieve polynomial bounds.
result Achieved polynomial bounds on previously exponential quantities, improving off-policy evaluation.
Adaptive reduction scheme approximates optimal policy in regularized MDPs.
problem Finding near optimal policy in regularized MDPs with biased solutions.
method Adaptive reduction of regularization parameter λ to approximate optimal policy.
result Iteration complexity reduced for obtaining ε-optimal policy.
RL struggles with generalization due to implicit partial observability.
problem Generalization in RL is difficult due to implicit partial observability.
method Re-cast RL problem as solving epistemic POMDPs and propose ensemble-based techniques.
result Simple ensemble-based technique achieves significant generalization gains.
We study Exo-MDPs to reduce sample complexity in reinforcement learning.
problem Reducing sample complexity in reinforcement learning for structured MDPs.
method Introducing Exo-MDPs and proving structural equivalence to linear mixture MDPs, establishing regret bounds.
result Proved O ( H 3 / 2 d K ) O(H^{3/2}d\sqrt{K}) O ( H 3/2 d K ) regret bound for Exo-MDPs, matching lower bounds. In this paper, a sparse Markov decision process (MDP) with novel causal sparse Tsallis entropy regularization is proposed.The proposed policy regularization induces a sparse and multi-modal optimal policy distribution of a sparse MDP. The full mathematical analysis of the proposed sparse MDP is provided.We first analyz…
Deep active inference learns policies from sensory inputs.
problem Learning policies in partially observable domains.
method Optimizes expected free energy with a variational autoencoder.
result Comparable or better performance than deep Q-learning.
Paper establishes new lower bounds for MDPs with changing transition kernels.
problem Minimizing sample complexity and regret in non-stationary MDPs.
method Developed novel lower bounds and constructed hard MDPs.
result Proved Ω ( ( H 3 S A / ε 2 ) log ( 1 / δ ) ) Ω((H^3SA/ε^2)\log(1/δ)) Ω (( H 3 S A / ε 2 ) log ( 1/ δ )) sample complexity lower bound. We consider the Markov Decision Process (MDP) of selecting a subset of items at each step, termed the Select-MDP (S-MDP). The large state and action spaces of S-MDPs make them intractable to solve with typical reinforcement learning (RL) algorithms especially when the number of items is huge. In this paper, we present …
New UCB algorithm for learning PSRs with tractable computation and accuracy.
problem Learning predictive state representations in sequential decision-making problems.
method Proposes a novel UCB-type algorithm with a bonus term to estimate PSRs accurately and efficiently.
result First known UCB-type approach for PSRs with guaranteed model accuracy and computational tractability.
We consider large-scale Markov decision processes (MDPs) with parameter uncertainty, under the robust MDP paradigm. Previous studies showed that robust MDPs, based on a minimax approach to handle uncertainty, can be solved using dynamic programming for small to medium sized problems. However, due to the "curse of dimen…
Optimizes learning policies in MDPs with weakly communicating structure.
problem Learning optimal policies in weakly communicating MDPs with generative model.
method Span-based approach, reducing to discounted MDPs for analysis.
result First minimax optimal sample complexity bound for weakly communicating MDPs.
A Markov Decision Process (MDP) is a popular model for reinforcement learning. However, its commonly used assumption of stationary dynamics and rewards is too stringent and fails to hold in adversarial, nonstationary, or multi-agent problems. We study an episodic setting where the parameters of an MDP can differ across…
New method removes oracle and reduces memory usage for robust MDPs.
problem Applying robust MDPs in practice due to model estimation and oracle requirements.
method Transformed robust MDPs into an alternative form allowing stochastic gradient methods and model-free approach.
result Sample-efficient algorithm with lower storage requirement and no oracle.
DeepAveragers solves offline RL by solving derived MDPs from static data.
problem Offline reinforcement learning with limited data.
method Solves derived non-parametric MDPs (DAC-MDPs) using deep representations and costs for under-represented parts.
result The approach can lower-bound performance and scale to complex offline RL problems.
New RL method learns to skip states in linearly q π q^π q π -realizable MDPs, simplifying to linear MDPs.
problem Online RL in episodic MDPs with linearly q π q^π q π -realizable action-values. method Derives a novel algorithm that learns to skip states and applies a linear MDP algorithm.
result First polynomial-sample-complexity online RL algorithm for linearly q π q^π q π -realizable MDPs. Optimizes learning policies in average-reward MDPs with improved sample complexity.
problem Learning optimal policies in average-reward MDPs with limited samples.
method Reduces to discounted MDPs and uses improved bounds for variance parameters.
result Establishes minimax optimal sample complexity bound of O(SA(H/ε^2))
Efficiently plans large MDPs with weak function approximations.
problem Planning in large MDPs with limited function approximation capabilities.
method Uses linear value function approximation with weak requirements and a generative oracle.
result Produces almost-optimal actions for any state with polynomial computation time.
New algorithm tackles confounding in offline RL for partially observable MDPs.
problem Confounding in offline reinforcement learning for partially observable MDPs.
method P3O algorithm using proximal causal inference and pessimistic confidence regions.
result Achieves n − 1 / 2 n^{-1/2} n − 1/2 -suboptimality for offline RL in POMDPs with confounded data. Reward suffices for convex MDPs, expanding RL to new problems.
problem Capturing goals as convex functions of stationary distribution.
method Reformulated as a min-max game using Fenchel duality.
result Convex MDPs require non-stationary reward functions.
Practical reinforcement learning problems are often formulated as constrained Markov decision process (CMDP) problems, in which the agent has to maximize the expected return while satisfying a set of prescribed safety constraints. In this study, we propose a novel simulator-based method to approximately solve a CMDP pr…
The paper addresses statistical estimation in MDPs with confounders using instrumental variables.
problem Statistical estimation of value functions in MDPs with unobservable confounders.
method Two-stage estimator based on instrumental variables for confounded linear MDPs.
result Established statistical properties of the two-stage estimator, including error bounds and asymptotic normality.
Study on regret minimization in deterministic MDPs.
problem Minimizing regret in deterministic reinforcement learning.
method Logarithmic regret lower bounds, leveraging graph theory and cycles.
result Explicitly quantifies the fundamental limit of performance achievable by any learning algorithm.
Improved regret bound for MNL MDPs with variance-aware approach.
problem Optimal reinforcement learning for MNL MDPs with structured variance.
method Introducing a problem-dependent constant measuring average variance, proposing an algorithm with improved regret bound.
result Minimax optimal regret bound of O ( d H 2 σ ˉ T T ) O(dH^2\barσ_T\sqrt{T}) O ( d H 2 σ ˉ T T ) for structured MDPs. This work uses action equivariance to learn structured latent spaces for reinforcement learning.
problem Learning structured latent spaces for reinforcement learning.
method Introduced a contrastive loss function to enforce action equivariance on learned representations.
result Optimal policies in the abstract MDP can be successfully lifted to the original MDP.
New method tackles MDPs by learning normalized representations efficiently.
problem Curse of dimensionality in MDPs.
method Contrastive representation learning for linear MDPs.
result First practical method with strong theoretical guarantees and empirical performance.
The Markov assumption (MA) is fundamental to the empirical validity of reinforcement learning. In this paper, we propose a novel Forward-Backward Learning procedure to test MA in sequential decision making. The proposed test does not assume any parametric form on the joint distribution of the observed data and plays an…