Study shows reinforcement learning algorithm's performance depends on metric space's size.
problem Continuous state and action spaces with metrics.
method Refined analysis of Sinclair et al. (2019) algorithm.
result Regret scales with the zooming dimension, a new parameter capturing near optimal actions.
New IRL algorithm for continuous state spaces with formal guarantees.
problem Finding a reward function for expert behavior in continuous state spaces.
method Modeling the system using orthonormal functions and providing correctness proofs.
result Proof of correctness and formal guarantees on sample and time complexity.
Paper proposes Vertex Networks for reinforcement learning of control systems with safety guarantees.
problem Challenges in reinforcement learning with hard state and action constraints.
method Vertex Networks incorporate safety constraints into policy network architecture, ensuring safety during exploration.
result Proposed Vertex Networks outperform vanilla reinforcement learning in benchmark control tasks.
KCRL learns stable policies for nonlinear systems with formal guarantees.
problem Lack of stabilization guarantees in RL methods for safety-critical systems.
method KCRL uses Krasovskii's Lyapunov functions as a stability constraint and a primal-dual approach to learn stabilizing policies.
result KCRL guarantees learning a stabilizing policy in a finite number of interactions.
New algorithm FLUTE achieves uniform-PAC convergence in RL with linear approx.
problem RL with linear function approximation lacks uniform-PAC guarantees.
method FLUTE algorithm with minimax value function estimator and multi-level partition scheme.
result Uniform-PAC convergence to optimal policy with high probability.
Private RL algorithm with privacy guarantees for personalized medicine decisions.
problem Privacy-preserving reinforcement learning for personalized medicine decisions.
method Developed a private optimism-based RL algorithm using joint differential privacy (JDP).
result Achieved strong PAC and regret bounds with a privacy guarantee.
Safe reinforcement learning framework using optimal transport for robustness.
problem Robustness and safety in deep reinforcement learning with limited data assumptions.
method Optimal transport perturbations to construct worst-case virtual state transitions.
result Significantly improved safety at deployment time compared to standard methods.
The paper tackles robust reinforcement learning with performance guarantees.
problem Finding a robust policy for RMDP with state space uncertainties.
method Proposes RLSPI algorithm for learning optimal robust policy with performance bounds.
result Demonstrates the performance of RLSPI on standard benchmark problems.
New algorithm for reinforcement learning reduces complexity and guarantees convergence.
problem Reinforcement learning problems with convex occupancy measures.
method MD-CURL, inspired by mirror descent, uses non-standard regularization.
result Achieves convergence guarantees and simple closed-form solution.
The paper reviews MARL and its causal challenges, advocating for a 'causality first' approach.
problem Challenges in multi-agent reinforcement learning (MARL) and lack of theoretical guarantees for emergent behavior.
method Discussion of how causal methods can improve safety, interpretability, and robustness in MARL.
result Causal methods can provide strong theoretical guarantees for emergent behavior in MARL.
Improved theoretical guarantees for SBEED algorithm.
problem Theoretical analysis of SBEED algorithm's performance.
method Near-optimal performance guarantee based on function classes and distribution shift.
result Improved guarantees for SBEED in terms of horizon and sample size.
Meta-algorithm for efficient reinforcement learning from human preferences.
problem Learning from human preference comparisons in Markov decision processes.
method Randomized exploration and experimental design for batch comparison queries.
result Meta-algorithm achieves both regret and last-iterate guarantees with minimal preference queries.
This paper introduces a new metric, ULI, for RL that ensures both cumulative and instantaneous performance.
problem High-stakes applications require RL algorithms to avoid playing bad policies.
method Introduces uniform last-iterate (ULI) guarantee, a stronger metric capturing both cumulative and instantaneous performance.
result ULI directly implies near-optimal cumulative performance across various metrics, but not the other way around.
Proves efficient learning of hierarchical structure in meta-reinforcement learning.
problem Lack of provable guarantees for learning hierarchical structures in reinforcement learning.
method Analyzed HRL in meta-RL setting with tabular transition dynamics, providing diversity conditions and regret bounds.
result Sample-efficient recovery of natural hierarchical structure with provable guarantees.
The paper explores learning good policies from past data in large state spaces.
problem Learning good policies from historical data in large state spaces.
method Introduces expressivity assumptions and data coverage for function approximation and algorithmic design.
result A variety of algorithms and their guarantees are presented based on assumptions and desired complexity.
New approach connects reinforcement learning robustness and regularization.
problem Dealing with external uncertainty in reinforcement learning.
method Introducing Wasserstein distributionally robust MDPs and a new regularizer.
result Established a dual relation between robust MDPs and regularization.
Offline RL algorithms protect privacy while learning from sensitive data.
problem Learning sensitive data in financial, legal, and healthcare applications.
method Differential privacy guarantees for offline RL algorithms.
result Provable prevention of privacy risks with strong learning bounds.
The research proposes a stopping rule for reinforcement learning algorithms based on instance-dependent confidence.
problem Dramatic variation in convergence rates of reinforcement learning algorithms due to problem structure.
method Develops instance-dependent confidence regions and a data-dependent stopping rule for MDP policy evaluation and optimal value estimation.
result Proposes a stopping rule that adapts to the instance-specific difficulty of the problem, allowing for early termination.
Algorithm tackles constrained reinforcement learning with concave-convex and knapsack constraints.
problem Constrained episodic reinforcement learning with concave rewards and convex constraints.
method Modular analysis with strong theoretical guarantees for concave-convex and knapsack settings.
result Significantly outperforms existing approaches in constrained episodic environments.
State entropy regularization improves robustness in reinforcement learning, especially under structured perturbations.
problem Structured and spatially correlated perturbations in reinforcement learning.
method State entropy regularization, compared to policy entropy.
result State entropy regularization provides better robustness to structured and spatially correlated perturbations.
QTRAN++ improves MARL performance in complex environments.
problem Poor empirical performance of QTRAN in complex environments.
method Stabilizing training objective, removing role separation, and introducing a multi-head mixing network.
result QTRAN++ achieves state-of-the-art performance in the Starcraft Multi-Agent Challenge (SMAC).
New algorithm for reinforcement learning in uncertain environments with unknown thresholds.
problem Safety in reinforcement learning in unknown and uncertain environments.
method Growing-Window estimator sampling and Stochastic Pessimistic-Optimistic Thresholding (SPOT) algorithm.
result Achieves sublinear regret and constraint violation of i l d e O ( T ) ilde{\mathcal{O}}(\sqrt{T}) i l d e O ( T ) . A new framework improves reinforcement learning algorithms with policy guarantees.
problem Designing efficient and stable reinforcement learning algorithms.
method A general framework (FMA-PG) based on functional mirror ascent that constructs surrogate functions enabling policy improvement guarantees.
result The proposed framework enables policy improvement guarantees that hold regardless of policy parameterization, and recovers important heuristics.
New method provides scalable safety guarantees for RL agents.
problem Safe reinforcement learning in real-life scenarios.
method State-augmentation and shield design for probabilistic avoidance.
result Strict formal safety guarantees for RL agents, scalable and practical.
In this work, we provide theoretical guarantees for reward decomposition in deterministic MDPs. Reward decomposition is a special case of Hierarchical Reinforcement Learning, that allows one to learn many policies in parallel and combine them into a composite solution. Our approach builds on mapping this problem into a…
The paper shows how to stabilize off-policy reinforcement learning using specific state representations.
problem Stability issues in reinforcement learning with function approximation and off-policy learning.
method Formal analysis of representation learning schemes based on the transition matrix of a policy.
result Schur and orthogonal bases of the Krylov subspace provide stable representations for TD learning.
We consider undiscounted reinforcement learning in Markov decision processes (MDPs) where both the reward functions and the state-transition probabilities may vary (gradually or abruptly) over time. For this problem setting, we propose an algorithm and provide performance guarantees for the regret evaluated against the…
New RL algorithms improve control tasks with data reuse.
problem Real-world control requires performance guarantees and data efficiency.
method Generalized Policy Improvement combining on-policy guarantees and sample reuse.
result Extensive experimental analysis shows benefits of new algorithms.
There have been numerous advances in reinforcement learning, but the typically unconstrained exploration of the learning process prevents the adoption of these methods in many safety critical applications. Recent work in safe reinforcement learning uses idealized models to achieve their guarantees, but these models do …
Safe exploration method for RL under disturbance ensures safety with probabilistic guarantees.
problem Safe reinforcement learning in real environments with disturbance.
method Uses partial prior knowledge and conservative inputs to ensure state constraint satisfaction.
result Guaranteed safety with pre-specified probability in the presence of stochastic disturbance.
Paper tackles efficient IRL in offline settings with polynomial samples and runtime.
problem Efficiently learning reward functions from expert demonstrations in offline settings.
method Adapting the pessimism principle for offline RL, achieving strong guarantees.
result Achieved efficient IRL in offline and online settings with polynomial samples and runtime.
A new MARL framework for community-based cooperation with transfer and active exploration.
problem Flexible coordination patterns in multi-agent systems with community structures.
method Community-based multi-agent reinforcement learning with transfer and active exploration.
result Provably convergent actor-critic algorithms for structured information sharing and transfer learning.
DARLING tackles non-stationary RL with guarantees, improving dynamic regret.
problem Non-stationary reinforcement learning in unknown change points.
method Detection Augmented Reinforcement Learning (DARLING) for tabular and linear MDPs.
result DARLING matches minimax lower bounds in tabular and linear MDPs.
New offline RL method works with limited data and function approximators.
problem Sample efficiency with limited data and weak function approximators.
method Pessimistic algorithm based on version space formed by marginalized importance sampling (MIS), with gap assumption.
result Guarantees sample efficiency for simple algorithm under specific assumptions.
OPSRL algorithm reduces regret with few samples in reinforcement learning.
problem High regret in reinforcement learning with limited data.
method Optimistic Posterior Sampling (OPSRL) with logarithmic sample complexity.
result Guaranteed high-probability regret bound of O ~ ( H 3 S A T ) \widetilde{\mathcal{O}}(\sqrt{H^3SAT}) O ( H 3 S A T ) . New algorithms achieve uniform-PAC guarantees for RL with bounded eluder dimension.
problem Achieving strong performance guarantees in reinforcement learning.
method Proposes algorithms for nonlinear bandits and model-based episodic RL with a bounded eluder dimension.
result Achieves uniform-PAC sample complexity that matches state-of-the-art regret bounds or sample complexity guarantees.
Two new algorithms improve Q* approximation in batch RL with linear error propagation.
problem Improving Q* approximation in batch reinforcement learning.
method Two novel algorithms that estimate Bellman error directly, without quadratic dependence.
result Linear-in-horizon error propagation for batch RL algorithms.
Unified analysis of tree-based methods for online reinforcement learning.
problem Designing efficient algorithms for online reinforcement learning with low sample complexity, storage, and computational burden.
method Unified theoretical analysis of tree-based hierarchical partitioning methods for online reinforcement learning.
result Our algorithms provide guarantees that scale with respect to the 'zooming dimension', improving upon ambient dimension scaling.
Extended analysis of Q-learning's efficiency, matching optimal regret.
problem Theoretical guarantees of Q-learning's efficiency and optimal regret.
method Survey of related research, detailed proof reasoning.
result Q-learning with UCB exploration achieves sample efficiency matching optimal regret.
Dealing with high variance is a significant challenge in model-free reinforcement learning (RL). Existing methods are unreliable, exhibiting high variance in performance from run to run using different initializations/seeds. Focusing on problems arising in continuous control, we propose a functional regularization appr…
Paper tackles offline RL with limited target samples using domain adaptation.
problem Limited samples in target dataset degrade offline RL performance.
method Proposes a framework to balance target and source datasets with theoretical guarantees.
result Establishes performance bounds and optimal weight for offline RL.
New reinforcement learning bound improves generalization for sequential data.
problem Challenges in obtaining generalization guarantees for reinforcement learning due to sequential data.
method PAC-Bayesian reinforcement learning with consideration of Markov dependencies and mixing time.
result Demonstrated practical utility through PB-SAC, providing meaningful confidence certificates.
Paper analyzes finite-time guarantees for preference-based RL.
problem Understanding finite-time guarantees for preference-based RL.
method Combines dueling bandits and policy search to navigate state space.
result Identifies best policy up to accuracy ε with high probability.
RH-UCRL combines pessimism and optimism for robust RL.
problem Ensuring reliable performance in real-world RL tasks with worst-case scenarios.
method RH-UCRL is a model-based RL algorithm that optimizes between an agent and an adversary, distinguishing between epistemic and aleatoric uncertainty.
result RH-UCRL achieves near-optimal sample complexity guarantees and outperforms other robust RL algorithms in adversarial environments.
We propose a general framework for policy representation for reinforcement learning tasks. This framework involves finding a low-dimensional embedding of the policy on a reproducing kernel Hilbert space (RKHS). The usage of RKHS based methods allows us to derive strong theoretical guarantees on the expected return of t…
New algorithms estimate Q-functions under partial coverage and realizability, improving offline RL guarantees.
problem Offline RL with limited exploration and assumptions about data coverage and Q-function realizability.
method Proposes minimax learning algorithms to estimate soft or vanilla Q-functions with L 2 L^2 L 2 -convergence guarantees. result PAC guarantees for offline RL under partial coverage and realizability conditions.
Study non-asymptotic BPI guarantees for online RL.
problem Identify optimal policy in MDP with high confidence.
method Non-asymptotic sample complexity guarantees for NaS algorithm.
result Sample complexity depends on MDP connectivity and curvature.
Reinforcement learning is a powerful paradigm for learning optimal policies from experimental data. However, to find optimal policies, most reinforcement learning algorithms explore all possible actions, which may be harmful for real-world systems. As a consequence, learning algorithms are rarely applied on safety-crit…