BriefGPT.xyz
May, 2017
Atari大挑战数据集
The Atari Grand Challenge Dataset
HTML
PDF
Vitaly Kurin, Sebastian Nowozin, Katja Hofmann, Lucas Beyer, Bastian Leibe
TL;DR
本文提出了一种减少数据使用量的方法,即利用人类示范数据对强化学习模型进行训练,在此基础上,作者基于Atari 2600回放数据集,发现示范数据的质量和模型的模仿学习性能之间有着密切关联,为进一步拓展该方法提供了研究方向。
Abstract
Recent progress in
reinforcement learning
(RL), fueled by its combination, with
deep learning
has enabled impressive results in learning to interact with complex virtual environments, yet real-world applications
→