BriefGPT.xyz
Dec, 2018
从单次演示中学习 Montezuma's Revenge
Learning Montezuma's Revenge from a Single Demonstration
HTML
PDF
Tim Salimans, Richard Chen
TL;DR
提出了一种新的利用单一示范来学习解决Montezuma's Revenge等复杂探索任务的方法，该方法通过最大化奖励来训练代理，缩短了学习时间，降低了任务复杂度。
Abstract
We propose a new method for learning from a
single demonstration
to solve hard
exploration tasks
like the Atari game
montezuma's revenge
.
→