paper-with-me

홈 › Papers

Safe Policy Improvement with Baseline Bootstrapping

2017-12-19 · Romain Laroche, Paul Trichelair, Rémi Tachet des Combes

This paper considers Safe Policy Improvement (SPI) in Batch Reinforcement Learning (Batch RL): from a fixed dataset and without direct access to the true environment, train a policy that is guaranteed to perform at least as well as the baseline policy used to collect the data. Our approach, called SPI with Baseline Bootstrapping (SPIBB), is inspired by the knows-what-it-knows paradigm: it bootstraps the trained policy with the baseline when the uncertainty is high. Our first algorithm, $\Pi_b$-SPIBB, comes with SPI theoretical guarantees. We also implement a variant, $\Pi_{\leq b}$-SPIBB, that is even more efficient in practice. We apply our algorithms to a motivational stochastic gridworld domain and further demonstrate on randomly generated MDPs the superiority of SPIBB with respect to existing algorithms, not only in safety but also in mean performance. Finally, we implement a model-free version of SPIBB and show its benefits on a navigation task with deep RL implementation called SPIBB-DQN, which is, to the best of our knowledge, the first RL algorithm relying on a neural network representation able to train efficiently and reliably from batch data, without any interaction with the environment.

📄 PDF Abstract BibTeX arXiv:1712.06924

Code (2)

RomainLaroche/SPIBB 공식 구현
rems75/SPIBB-DQN 공식 구현 pytorch

Tasks

Reinforcement Learning

Similar Papers 제목 키워드 기반

Safe Policy Improvement with Soft Baseline Bootstrapping

2019-07-11 · Kimia Nadjahi, Romain Laroche, Rémi Tachet des Combes

Batch Reinforcement Learning (Batch RL) consists in training a policy using trajectories collected with another policy, called the behavioural policy. Safe policy improvement (SPI) provides guarantees with high probabili…

Reinforcement Learning

Safe Policy Improvement with an Estimated Baseline Policy

2019-09-11 · Thiago D. Simão, Romain Laroche, Rémi Tachet des Combes

Previous work has shown the unreliability of existing algorithms in the batch Reinforcement Learning setting, and proposed the theoretically-grounded Safe Policy Improvement with Baseline Bootstrapping (SPIBB) fix: repro…

ManagementReinforcement Learning

Safe Policy Improvement Approaches and their Limitations

2022-08-01 · Philipp Scholl, Felix Dietrich, Clemens Otte, Steffen Udluft

Safe Policy Improvement (SPI) is an important technique for offline reinforcement learning in safety critical applications as it improves the behavior policy with a high probability. We classify various SPI approaches fr…

Safe Policy Improvement Approaches on Discrete Markov Decision Processes

2022-01-28 · Philipp Scholl, Felix Dietrich, Clemens Otte, Steffen Udluft

Safe Policy Improvement (SPI) aims at provable guarantees that a learned policy is at least approximately as good as a given baseline policy. Building on SPI with Soft Baseline Bootstrapping (Soft-SPIBB) by Nadjahi et al…

Multi-Objective SPIBB: Seldonian Offline Policy Improvement with Safety Constraints in Finite MDPs

2021-05-31 · NeurIPS 2021 12 · Harsh Satija, Philip S. Thomas, Joelle Pineau, Romain Laroche

We study the problem of Safe Policy Improvement (SPI) under constraints in the offline Reinforcement Learning (RL) setting. We consider the scenario where: (i) we have a dataset collected under a known baseline policy, (…

Reinforcement Learning (RL)