arXivDaily arXiv每日学术速递 周一至周五更新
arXiv周末暂无论文更新,休息一下吧,周末愉快~~
arXiv 2008.03273cs.LGcs.SYeess.SYstat.ML
收藏

SafePILCO:一种安全且数据高效的策略合成软件工具

SafePILCO: a software tool for safe and data-efficient policy synthesis

University of Oxford(牛津大学)

Kyriakos Polymenakos, Nikitas Rontsis, Alessandro Abate, Stephen Roberts

摘要:

SafePILCO是一个软件工具,用于通过强化学习进行安全和数据高效的策略搜索。它扩展了最初用MATLAB编写的已知PILCO算法,以支持安全学习。我们提供了一个Python实现,并利用现有的库,使代码库保持简短和模块化,这适合验证,强化学习和控制社区的更广泛使用。

英文摘要:

SafePILCO is a software tool for safe and data-efficient policy search with reinforcement learning. It extends the known PILCO algorithm, originally written in MATLAB, to support safe learning. We provide a Python implementation and leverage existing libraries that allow the codebase to remain short and modular, which is appropriate for wider use by the verification, reinforcement learning, and control communities.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑