强化学习资料

学习资料重要

相关博客：http://blog.csdn.net/dark_scope/article/details/8252969

专栏：http://blog.csdn.net/column/details/deeprl.html

增强学习课程 David Silver （有视频和ppt）:http://www0.cs.ucl.ac.uk/staff/D.Silver/web/Teaching.html

最好的增强学习教材：

Reinforcement Learning: An Introduction：https://webdocs.cs.ualberta.ca/~sutton/book/the-book.html

深度学习课程 （有视频有ppt有作业）：Machine Learning

深度增强学习的讲座都是David Silver的：

ICLR 2015 part 1 https://www.youtube.com/watch?v=EX1CIVVkWdE

ICLR 2015 part 2 https://www.youtube.com/watch?v=zXa6UFLQCtg

UAI 2015 https://www.youtube.com/watch?v=qLaDWKd61Ig

RLDM 2015 Deep Reinforcement Learning - VideoLectures.NET

其他课程：

增强学习

Michael Littman: https://www.udacity.com/course/reinforcement-learning–ud600

AI（包含增强学习，使用Pacman实验）

Pieter Abbeel：ColumbiaX: Artificial Intelligence (AI) | edX

Deep reinforcement Learning:

Pieter Abbeel  http://rll.berkeley.edu/deeprlcourse/

高级机器人技术（Advanced Robotics）：

Pieter Abbeel：http://www.cs.berkeley.edu/~pabbeel/cs287-fa15/

深度学习相关课程：

用于视觉识别的卷积神经网络（Convolutional Neural Network for visual network):CS231n Convolutional Neural Networks for Visual Recognition

机器学习 Machine Learning

Andrew Ng:

Supervised Machine Learning: Regression and Classification | Coursera

CS229: Machine Learning

神经网络（Neural Network for Machine Learning)(2012年的）

Hinton:https://www.coursera.org/course/neuralnets

最新机器人专题课程Penn（2016年开课）：Robotics Specialization [6 courses] (Penn) | Coursera

2 论文资料

GitHub - junhyukoh/deep-reinforcement-learning-papers: A list of recent papers regarding deep reinforcement learning

GitHub - muupan/deep-reinforcement-learning-papers: A list of papers and resources dedicated to deep reinforcement learning

这两个人收集的基本涵盖了当前deep reinforcement learning 的论文资料。

3 大牛情况：

DeepMind：http://www.deepmind.com/publications.html

Pieter Abbeel 团队：http://www.eecs.berkeley.edu/~pabbeel/

Satinder Singh:Home page for Satinder Singh (Baveja) and Reinforcement Learning

CMU 进展：Lerrel Pinto

Prefered Networks: (日本创业公司）

4 会议情况

Deep Reinforcement Learning Workshop NIPS 2015 : Deep Reinforcement Learning Workshop

深度学习研究总结：强化学习技术趋势与分析（经典论文）

ICLR 2017中和Deep Reinforcement Learning相关的论文我这边收集了一下，一共有30篇（可能有漏），大部分来自于DeepMind和OpenAI，可见DRL依然主要由DeepMind和OpenAI把持。

2 DeepMind的论文分析

[1] LEARNING TO COMPOSE WORDS INTO SENTENCES WITH REINFORCEMENT LEARNING

[2] LEARNING TO NAVIGATE IN COMPLEX ENVIRONMENTS

[3] LEARNING TO PERFORM PHYSICS EXPERIMENTS VIA DEEP REINFORCEMENT LEARNING

[4] PGQ: COMBINING POLICY GRADIENT AND Q- LEARNING

[5] Q-PROP: SAMPLE-EFFICIENT POLICY GRADIENT WITH AN OFF-POLICY CRITIC

[6] REINFORCEMENT LEARNING WITH UNSUPERVISED AUXILIARY TASKS

[7] SAMPLE EFFICIENT ACTOR-CRITIC WITH EXPERIENCE REPLAY

[8] THE PREDICTRON: END-TO-END LEARNING AND PLANNING

3 OpenAI的论文分析（包含Sergey Levine的论文）

[9] #EXPLORATION: A STUDY OF COUNT-BASED EXPLORATION FOR DEEP REINFORCEMENT LEARNING

[10] GENERALIZING SKILLS WITH SEMI-SUPERVISED REINFORCEMENT LEARNING

[11] LEARNING INVARIANT FEATURE SPACES TO TRANS- FER SKILLS WITH REINFORCEMENT LEARNING

[12] LEARNING VISUAL SERVOING WITH DEEP FEATURES AND TRUST REGION FITTED Q-ITERATION

[13] MODULAR MULTITASK REINFORCEMENT LEARNING WITH POLICY SKETCHES

[14] STOCHASTIC NEURAL NETWORKS FOR HIERARCHICAL REINFORCEMENT LEARNING

[15] THIRD PERSON IMITATION LEARNING

[16] UNSUPERVISED PERCEPTUAL REWARDS FOR IMITATION LEARNING

[17] EPOPT: LEARNING ROBUST NEURAL NETWORK POLICIES USING MODEL ENSEMBLES 

[18] RL2: FAST REINFORCEMENT LEARNING VIA SLOW REINFORCEMENT LEARNING

4 其他论文

[19] COMBATING DEEP REINFORCEMENT LEARNING’S SISYPHEAN CURSE WITH INTRINSIC FEAR

[20] COMMUNICATING HIERARCHICAL NEURAL CONTROLLERS FOR LEARNING
ZERO-SHOT TASK GENERALIZATION

[21] DESIGNING NEURAL NETWORK ARCHITECTURES USING REINFORCEMENT LEARNING

[22] LEARNING TO PLAY IN A DAY: FASTER DEEP REIN- FORCEMENT LEARNING BY OPTIMALITY TIGHTENING

[23] LEARNING TO REPEAT: FINE GRAINED ACTION REPETITION FOR DEEP REINFORCEMENT LEARNING

[24] MULTI-TASK LEARNING WITH DEEP MODEL BASED REINFORCEMENT LEARNING

[25] NEURAL ARCHITECTURE SEARCH WITH REINFORCEMENT LEARNING

[26] OPTIONS DISCOVERY WITH BUDGETED REINFORCE- MENT LEARNING

[27] REINFORCEMENT LEARNING THROUGH ASYNCHRONOUS ADVANTAGE ACTOR-CRITIC ON A GPU

[28] SPATIO-TEMPORAL ABSTRACTIONS IN REINFORCEMENT LEARNING THROUGH NEURAL ENCODING

[29] SURPRISE-BASED INTRINSIC MOTIVATION FOR DEEP REINFORCEMENT LEARNING

[30] TUNING RECURRENT NEURAL NETWORKS WITH REINFORCEMENT LEARNING