- 1、本文档共16页,可阅读全部内容。
- 2、有哪些信誉好的足球投注网站(book118)网站文档一经付费(服务费),不意味着购买了该文档的版权,仅供个人/单位学习、研究之用,不得用于商业用途,未经授权,严禁复制、发行、汇编、翻译或者网络传播等,侵权必究。
- 3、本站所有内容均由合作方或网友上传,本站不对文档的完整性、权威性及其观点立场正确性做任何保证或承诺!文档内容仅供研究参考,付费前请自行鉴别。如您付费,意味着您自己接受本站规则且自行承担风险,本站不退款、不进行额外附加服务;查看《如何避免下载的几个坑》。如果您已付费下载过本站文档,您可以点击 这里二次下载。
- 4、如文档侵犯商业秘密、侵犯著作权、侵犯人身权等,请点击“版权申诉”(推荐),也可以打举报电话:400-050-0827(电话支持时间:9:00-18:30)。
- 5、该文档为VIP文档,如果想要下载,成为VIP会员后,下载免费。
- 6、成为VIP后,下载本文档将扣除1次下载权益。下载后,不支持退款、换文档。如有疑问请联系我们。
- 7、成为VIP后,您将拥有八大权益,权益包括:VIP文档下载权益、阅读免打扰、文档格式转换、高级专利检索、专属身份标志、高级客服、多端互通、版权登记。
- 8、VIP文档为合作方或网友上传,每下载1次, 网站将根据用户上传文档的质量评分、类型等,对文档贡献者给予高额补贴、流量扶持。如果你也想贡献VIP文档。上传文档
查看更多
Published as a conference paper at ICLR 2016
ACTOR-MIMIC
DEEP MULTITASK AND TRANSFER REINFORCEMENT
LEARNING
Emilio Parisotto, Jimmy Ba, Ruslan Salakhutdinov
Department of Computer Science
University of Toronto
Toronto, Ontario, Canada
eparisotto,jimmy,rsalakhu@cs.toronto.edu
6
1
0
2 ABSTRACT
b
e The ability to act in multiple environments and transfer previous knowledge to
F new situations can be considered a critical aspect of any intelligent agent. To-
2 wards this goal, we define a novel method of multitask and transfer learning that
2 enables an autonomous agent to learn how to behave in multiple tasks simultane-
ously, and then generalize its knowledge to new domains. This method, termed
]
G “Actor-Mimic”, exploits the use of deep reinforcement learning and model com-
pression techniques to train a single policy network that learns how to act in a set
L of distinct tasks by using the guidance of several expert teachers. We then show
.
s that the representations learnt by the deep policy network are capable of general-
c izing to new tasks with no prior expert guidance, speeding up learning in novel
[
environments. Although our method can in general be applied to a wide range
4 of problems, we use Atari games as a testing environment to demonstrate these
v
您可能关注的文档
- 浅谈信息学竞赛中的“0”和“1”.ppt
- 深度学习论文Matching Networks for One Shot Learning_20180118194149.pdf
- 深度学习论文Progressive Neural Networks_20180118194149.pdf
- 深度学习论文POLICY DISTILLATION_20180118194149.pdf
- 深度学习论文Lifelong Machine Learning Systems Beyond Learning Algorithms_20180118194149.pdf
- 深度学习论文Deep Learning of Representations for Unsupervised and_20180118194149.pdf
- 深度学习论文Trust Region Policy Optimization_20180118194150.pdf
- 深度学习论文Mastering the game of Go with deep_20180118194149.pdf
- 深度学习论文Dueling Network Architectures for Deep Reinforcement Learning.pdf
- 深度学习论文Continuous Deep Q-Learning with Model-based Acceleration_20180118194149.pdf
文档评论(0)