BriefGPT.xyz
Aug, 2020
一种适用于强化学习任务的组合规格语言
A Composable Specification Language for Reinforcement Learning Tasks
HTML
PDF
Kishor Jothimurugan, Rajeev Alur, Osbert Bastani
TL;DR
提出一种用于规定复杂控制任务的语言并实现在工具SPECTRL中的算法，可以将任务规定编译为奖励函数并自动进行奖励形状。
Abstract
reinforcement learning
is a promising approach for learning
control policies
for
robot tasks
. However, specifying complex tasks (e.g., wit
→