【AlphaGoZero核心技术】深度强化学习知识资料全集(论文/代码/教程/视频/文章等)- 专知

会员服务 ·

【AlphaGoZero核心技术】深度强化学习知识资料全集(论文/代码/教程/视频/文章等)

钱胜胜/Quan

点击上方“专知”关注获取更多AI知识!

【导读】昨天 Google DeepMind在Nature上发表最新论文，介绍了迄今最强最新的版本AlphaGo Zero，不使用人类先验知识，使用纯强化学习，将价值网络和策略网络整合为一个架构，3天训练后就以100比0击败了上一版本的AlphaGo。Alpha Zero的背后核心技术是深度强化学习，为此，专知特别收录整理聚合了关于强化学习的最全知识资料，欢迎大家查看！

先看下Google DeepMind 研究人员David Silver介绍 AlphaGo Zero：

专知 -Deep Reinforcement Learning 最全资料集合：

Nature 论文
Mastering the game of Go without human knowledge
Nature 550, 7676 (2017). doi:10.1038/nature24270
Authors: David Silver, Julian Schrittwieser, Karen Simonyan, Ioannis Antonoglou, Aja Huang, Arthur Guez, Thomas Hubert, Lucas Baker, Matthew Lai, Adrian Bolton, Yutian Chen, Timothy Lillicrap, Fan Hui, Laurent Sifre, George van den Driessche, Thore Graepel & Demis Hassabis
网址：https://www.nature.com/nature/journal/v550/n7676/full/nature24270.html
请下载pdf查看！

Mastering the game of Go with deep neural networks and tree search

David Silver, Aja Huang, Chris J. Maddison, Arthur Guez, Laurent Sifre, George van den Driessche, Julian Schrittwieser, Ioannis Antonoglou, Vedavyas Panneershelvam, Marc Lanctot, Sander Dieleman, Dominik Grewe, John Nham, Nal Kalchbrenner, Ilya Sutskever, Timothy P. Lillicrap, Madeleine Leach, Koray Kavukcuoglu, Thore Graepel, Demis Hassabis:
Nature 529(7587): 484-489 (2016)

Papers

Mastering the Game of Go without Human Knowledge	https://deepmind.com/documents/119/agz_unformatted_nature.pdf
Human level control with deep reinforcement learning	http://www.nature.com/nature/journal/v518/n7540/full/nature14236.html
Play Atari game with deep reinforcement learning	https://www.cs.toronto.edu/%7Evmnih/docs/dqn.pdf
Prioritized experience replay	https://arxiv.org/pdf/1511.05952v2.pdf
Dueling DQN	https://arxiv.org/pdf/1511.06581v3.pdf
Deep reinforcement learning with double Q Learning	https://arxiv.org/abs/1509.06461
Deep Q learning with NAF	https://arxiv.org/pdf/1603.00748v1.pdf
Deterministic policy gradient	http://jmlr.org/proceedings/papers/v32/silver14.pdf
Continuous control with deep reinforcement learning) (DDPG)	https://arxiv.org/pdf/1509.02971v5.pdf
Asynchronous Methods for Deep Reinforcement Learning	https://arxiv.org/abs/1602.01783
Policy distillation	https://arxiv.org/abs/1511.06295
Control of Memory, Active Perception, and Action in Minecraft	https://arxiv.org/pdf/1605.09128v1.pdf
Unifying Count-Based Exploration and Intrinsic Motivation	https://arxiv.org/pdf/1606.01868v2.pdf
Incentivizing Exploration In Reinforcement Learning With Deep Predictive Models	https://arxiv.org/pdf/1507.00814v3.pdf
Action-Conditional Video Prediction using Deep Networks in Atari Games	https://arxiv.org/pdf/1507.08750v2.pdf
Control of Memory, Active Perception, and Action in Minecraft	https://web.eecs.umich.edu/~baveja/Papers/ICML2016.pdf
PathNet	https://arxiv.org/pdf/1701.08734.pdf

Papers for NLP

Coarse-to-Fine Question Answering for Long Documents	https://homes.cs.washington.edu/~eunsol/papers/acl17eunsol.pdf
A Deep Reinforced Model for Abstractive Summarization	https://arxiv.org/pdf/1705.04304.pdf
Reinforcement Learning for Simultaneous Machine Translation	https://www.umiacs.umd.edu/~jbg/docs/2014_emnlp_simtrans.pdf
Dual Learning for Machine Translation	https://papers.nips.cc/paper/6469-dual-learning-for-machine-translation.pdf
Learning to Win by Reading Manuals in a Monte-Carlo Framework	http://people.csail.mit.edu/regina/my_papers/civ11.pdf
Improving Information Extraction by Acquiring External Evidence with Reinforcement Learning	http://people.csail.mit.edu/regina/my_papers/civ11.pdf
Deep Reinforcement Learning with a Natural Language Action Space	http://www.aclweb.org/anthology/P16-1153
Deep Reinforcement Learning for Dialogue Generation	https://arxiv.org/pdf/1606.01541.pdf
Reinforcement Learning for Mapping Instructions to Actions	http://people.csail.mit.edu/branavan/papers/acl2009.pdf
Language Understanding for Text-based Games using Deep Reinforcement Learning	https://arxiv.org/pdf/1506.08941.pdf
End-to-end LSTM-based dialog control optimized with supervised and reinforcement learning	https://arxiv.org/pdf/1606.01269v1.pdf
End-to-End Reinforcement Learning of Dialogue Agents for Information Access	https://arxiv.org/pdf/1609.00777v1.pdf
Hybrid Code Networks: practical and efficient end-to-end dialog control with supervised and reinforcement learning	https://arxiv.org/pdf/1702.03274.pdf
Deep Reinforcement Learning for Mention-Ranking Coreference Models	https://arxiv.org/abs/1609.08667

精选文章

wiki	https://en.wikipedia.org/wiki/Reinforcement_learning
Deep Reinforcement Learning: Pong from Pixels	http://karpathy.github.io/2016/05/31/rl/
CS 294: Deep Reinforcement Learning	http://rll.berkeley.edu/deeprlcourse/
强化学习系列之一:马尔科夫决策过程	http://www.algorithmdog.com/%E5%BC%BA%E5%8C%96%E5%AD%A6%E4%B9%A0-%E9%A9%AC%E5%B0%94%E7%A7%91%E5%A4%AB%E5%86%B3%E7%AD%96%E8%BF%87%E7%A8%8B
强化学习系列之九:Deep Q Network (DQN)	http://www.algorithmdog.com/drl
强化学习系列之三:模型无关的策略评价	http://www.algorithmdog.com/reinforcement-learning-model-free-evalution
【整理】强化学习与MDP	http://www.cnblogs.com/mo-wang/p/4910855.html
强化学习入门及其实现代码	http://www.jianshu.com/p/165607eaa4f9
深度强化学习系列（二）：强化学习	http://blog.csdn.net/ikerpeng/article/details/53031551
采用深度 Q 网络的 Atari 的 Demo： Nature 上关于深度 Q 网络 (DQN) 论文:	http://www.nature.com/articles/nature14236
David视频里所使用的讲义pdf	https://pan.baidu.com/s/1nvqP7dB
什么是强化学习？	http://www.cnblogs.com/geniferology/p/what_is_reinforcement_learning.html
DavidSilver 关于深度确定策略梯度 DPG的论文	http://www.jmlr.org/proceedings/papers/v32/silver14.pdf
Nature 上关于 AlphaGo 的论文：	http://www.nature.com/articles/nature16961
AlphaGo 相关的资源	deepmind.com/research/alphago/
What’s the Difference Between Artificial Intelligence, Machine Learning, and Deep Learning?	https://blogs.nvidia.com/blog/2016/07/29/whats-difference-artificial-intelligence-machine-learning-deep-learning-ai/
Deep Learning in a Nutshell: Reinforcement Learning	https://devblogs.nvidia.com/parallelforall/deep-learning-nutshell-reinforcement-learning/
Bellman equation	https://en.wikipedia.org/wiki/Bellman_equation
Reinforcement learning	https://en.wikipedia.org/wiki/Reinforcement_learning
Mastering the Game of Go without Human Knowledge	https://deepmind.com/documents/119/agz_unformatted_nature.pdf
Reinforcement Learning(RL) for Natural Language Processing(NLP)	https://github.com/adityathakker/awesome-rl-nlp

视频教程

强化学习教程(莫烦)	https://morvanzhou.github.io/tutorials/machine-learning/reinforcement-learning/
强化学习课程 by David Silver	https://www.bilibili.com/video/av8912293/?from=search&seid=1166472326542614796
CS234: Reinforcement Learning	http://web.stanford.edu/class/cs234/index.html
什么是强化学习? (Reinforcement Learning)	https://www.youtube.com/watch?v=NVWBs7b3oGk
什么是 Q Learning (Reinforcement Learning 强化学习)	https://www.youtube.com/watch?v=HTZ5xn12AL4
强化学习-莫烦	https://morvanzhou.github.io/tutorials/machine-learning/ML-intro/
David Silver深度强化学习第1课 - 简介 (中文字幕)	https://www.bilibili.com/video/av9831889/
David Silver的这套视频公开课（Youtube）	https://www.youtube.com/watch?v=2pWv7GOvuf0&amp;amp;amp;amp;list=PL7-jPKtc4r78-wCZcQn5IqyuWhBZ8fOxT
David Silver的这套视频公开课（Bilibili）	http://www.bilibili.com/video/av9831889/?from=search&seid=17387316110198388304
Deep Reinforcement Learning	http://videolectures.net/rldm2015_silver_reinforcement_learning/

Tutorial

Reinforcement Learning for NLP	http://www.umiacs.umd.edu/~jbg/teaching/CSCI_7000/11a.pdf
ICML 2016, Deep Reinforcement Learning tutorial	http://icml.cc/2016/tutorials/deep_rl_tutorial.pdf
DQN tutorial	https://medium.com/@awjuliani/simple-reinforcement-learning-with-tensorflow-part-4-deep-q-networks-and-beyond-8438a3e2b8df#.28wv34w3a

代码

OpenAI Gym	https://github.com/openai/gym
GoogleDeep Mind 团队深度 Q 网络 (DQN) 源码:	http://sites.google.com/a/deepmind.com/dqn/
ReinforcementLearningCode	https://github.com/halleanwoo/ReinforcementLearningCode
reinforcement-learning	https://github.com/dennybritz/reinforcement-learning
DQN	https://github.com/devsisters/DQN-tensorflow
DDPG	https://github.com/stevenpjg/ddpg-aigym
A3C01	https://github.com/miyosuda/async_deep_reinforce
A3C02	https://github.com/openai/universe-starter-agent