paper-with-me

홈 › Papers

Benchmarks for Reinforcement Learning with Biased Offline Data and Imperfect Simulators

2024-06-30 · Ori Linial, Guy Tennenholtz, Uri Shalit

In many reinforcement learning (RL) applications one cannot easily let the agent act in the world; this is true for autonomous vehicles, healthcare applications, and even some recommender systems, to name a few examples. Offline RL provides a way to train agents without real-world exploration, but is often faced with biases due to data distribution shifts, limited coverage, and incomplete representation of the environment. To address these issues, practical applications have tried to combine simulators with grounded offline data, using so-called hybrid methods. However, constructing a reliable simulator is in itself often challenging due to intricate system complexities as well as missing or incomplete information. In this work, we outline four principal challenges for combining offline data with imperfect simulators in RL: simulator modeling error, partial observability, state and action discrepancies, and hidden confounding. To help drive the RL community to pursue these problems, we construct ``Benchmarks for Mechanistic Offline Reinforcement Learning'' (B4MRL), which provide dataset-simulator benchmarks for the aforementioned challenges. Our results suggest the key necessity of such benchmarks for future research.

📄 PDF Abstract BibTeX arXiv:2407.00806

Code (0)

등록된 구현이 없습니다.

Tasks

Autonomous VehiclesOffline RLRecommendation Systemsreinforcement-learningReinforcement LearningReinforcement Learning (RL)

Similar Papers 제목 키워드 기반

Survival Instinct in Offline Reinforcement Learning

2023-06-05 · NeurIPS 2023 11

We present a novel observation about the behavior of offline reinforcement learning (RL) algorithms: on many benchmark datasets, offline RL can produce well-performing and safe policies even when trained with "wrong" rew…

Offline RLreinforcement-learningReinforcement LearningReinforcement Learning (RL)

Mind the Gap: Offline Policy Optimization for Imperfect Rewards

2023-02-03 · Jianxiong Li, Xiao Hu, Haoran Xu, Jingjing Liu 외

Reward function is essential in reinforcement learning (RL), serving as the guiding signal to incentivize agents to solve given tasks, however, is also notoriously difficult to design. In many cases, only imperfect rewar…

Reinforcement Learning (RL)

Data-Augmented Game Starts for Accelerating Self-Play Exploration in Imperfect Information Games

2026-05-14 · JB Lanier, Nathan Monette, Pierre Baldi, Roy Fox arxiv

Finding approximate equilibria for large-scale imperfect-information competitive games such as StarCraft, Dota, and CounterStrike remains computationally infeasible due to sparse rewards and challenging exploration over …

Reinforcement Learning

OASIS: Conditional Distribution Shaping for Offline Safe Reinforcement Learning

2024-07-19 · Yihang Yao, Zhepeng Cen, Wenhao Ding, Haohong Lin 외

Offline safe reinforcement learning (RL) aims to train a policy that satisfies constraints using a pre-collected dataset. Most current methods struggle with the mismatch between imperfect demonstrations and the desired s…

reinforcement-learningReinforcement LearningReinforcement Learning (RL)Safe Reinforcement Learning

Counterfactual-Augmented Importance Sampling for Semi-Offline Policy Evaluation

2023-10-26 · NeurIPS 2023 11 · Shengpu Tang, Jenna Wiens

In applying reinforcement learning (RL) to high-stakes domains, quantitative and qualitative evaluation using observational data can help practitioners understand the generalization performance of new policies. However, …

counterfactualOff-policy evaluationReinforcement Learning (RL)