BriefGPT.xyz
Jun, 2019
基于好奇心的多标准事后经验回放
Curiosity-Driven Multi-Criteria Hindsight Experience Replay
HTML
PDF
John B. Lanier, Stephen McAleer, Pierre Baldi
TL;DR
本文提出一种方法，将后见之明与好奇心驱动探索和课程学习相结合，以解决具有挑战性的稀疏奖励堆叠块任务，并且此方法成功地实现了在机器人手臂上堆叠两个以上的块，而无须使用人的演示。
Abstract
Dealing with
sparse rewards
is a longstanding challenge in reinforcement learning. The recent use of
hindsight methods
have achieved success on a variety of sparse-reward tasks, but they fail on complex tasks suc
→