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 2504.03420cs.LGcs.AIcs.CV
收藏

面向Deep-RL稀疏奖励场景的自主状态空间分割

Autonomous state-space segmentation for Deep-RL sparse reward scenarios

Institute of Cognitive Sciences and Technologies (ISTC) National Council of Research (CNR)(国家研究委员会认知科学与技术研究所(ISTC))

Gianluca Maselli, Vieri Giuliano Santucci

AI总结:

针对稀疏奖励下的Deep Reinforcement Learning问题,提出一种双层架构交替进行内在驱动探索与目标导向策略学习,通过构建专门子网络避免重复探索,在Gym SuperMarioBros中验证了自主环境分割生成高效路径的有效性。

摘要:

处理稀疏奖励环境对于旨在自主开放式学习环境中运行的系统一直至关重要。内在动机是帮助Deep Reinforcement Learning算法在此类场景中学习的有效方法。事实上,当外在奖励延迟或缺失时,通常采用新颖性或好奇心等内在奖励信号来改善探索。基于先前的工作,我们通过提出一种双层架构来解决稀疏奖励下的策略学习问题,该架构在“内在驱动”的探索与自主子目标生成阶段,和稀疏奖励、目标导向的策略学习阶段之间交替。其核心思想是构建多个小型网络,每个网络专门负责特定子路径,并将其作为未来探索的起点,无需从头探索已学过的路径。系统的两个版本在Gym SuperMarioBros环境中进行了训练和测试,未考虑任何额外外在奖励。结果表明了该方法的有效性,以及自主分割环境以生成通向最终目标高效路径的重要性。

英文摘要:

Dealing with environments with sparse rewards has always been crucial for systems developed to operate in autonomous open-ended learning settings. Intrinsic Motivations could be an effective way to help Deep Reinforcement Learning algorithms learn in such scenarios. In fact, intrinsic reward signals, such as novelty or curiosity, are generally adopted to improve exploration when extrinsic rewards are delayed or absent. Building on previous works, we tackle the problem of learning policies in the presence of sparse rewards by proposing a two-level architecture that alternates an ''intrinsically driven'' phase of exploration and autonomous sub-goal generation, to a phase of sparse reward, goal-directed policy learning. The idea is to build several small networks, each one specialized on a particular sub-path, and use them as starting points for future exploration without the need to further explore from scratch previously learnt paths. Two versions of the system have been trained and tested in the Gym SuperMarioBros environment without considering any additional extrinsic reward. The results show the validity of our approach and the importance of autonomously segment the environment to generate an efficient path towards the final goal.

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。

↑