BriefGPT.xyz
Jan, 2024
通过自动学习组合子任务实现高效样本强化学习
Sample Efficient Reinforcement Learning by Automatically Learning to Compose Subtasks
HTML
PDF
Shuai Han, Mehdi Dastani, Shihan Wang
TL;DR
自动结构化奖励函数以提高样本利用率,并在稀疏奖励环境中显著优于现有技术基线。
Abstract
Improving
sample efficiency
is central to
reinforcement learning
(RL), especially in environments where the rewards are sparse. Some recent approaches have proposed to specify reward functions as manually designe
→