Papers:
- NLP
DRL4NLP
- https://github.com/ganeshjawahar/drl4nlp.scratchpad
+++Policy Gradients buck_arxiv17: Ask the Right Questions: Active Question Reformulation with Reinforcement Learning [arXiv] dhingra_acl17: Towards End-to-End Reinforcement Learning of Dialogue Agents for Information Access [arXiv] [code] paulus_arxiv17: A Deep Reinforced Model for Abstractive Summarization [arXiv] nogueira_arxiv17: Task-Oriented Query Reformulation with Reinforcement Learning [arXiv] [code] li_iclr17: Dialog Learning with Human-in-the-loop [arXiv] [code] li_iclr17_2: Learning through dialogue interactions by asking questions [arXiv] [code] yogatama_iclr17: Learning to Compose Words into Sentences with Reinforcement Learning [arXiv] dinu_nips16w: Reinforcement Learning for Transition-Based Mention Detection [arXiv] clark_emnlp16: Deep Reinforcement Learning for Mention-Ranking Coreference models [arXiv] [code] +++Value Function narasimhan_emnlp16: Improving Information Extraction by Acquiring External Evidence with Reinforcement Learning [arXiv] [code] +++Misc bordes_iclr17: Learning End-to-End Goal-Oriented Dialog [arXiv] weston_nips16: Dialog-based Language Learning [arXiv] [code] nogueira_nips16: End-to-End Goal-Driven Web Navigation [arXiv] [code]
- https://github.com/ganeshjawahar/drl4nlp.scratchpad
Playing Atari with Deep Reinforcement Learning
- Deep Reinforcement Learning with a Natural Language Action Space
- https://arxiv.org/pdf/1511.04636.pdf
- Bidirectional LSTM-CRF Models for Sequence Tagging
- Hinton 2006 Deep Belief Network
- list:
- A fast learning algorithm for deep belief nets. Neural Computation
- Greedy Layer-Wise Training of Deep Networks
- Efficient Learning of Sparse Representations with an Energy-Based Model
- key point:
- unsup learn for pre train
- train by layers
- sup learn for tuning weight between layers
- list:
Office sites:
- Kaggle
- pygame
- Deep Reinforcement Learning for Keras.
- ai-code
- DeepMind
- 库
开始尝试机器学习库可以从安装最基础也是最重要的开始，像numpy和scipy。
- 查看和执行数据操作：pandas（http://pandas.pydata.org/）
- 对于各种机器学习模型：scikit-learn（http://scikit-learn.org/stable/）
- 最好的gradient boosting库：xgboost（https://github.com/dmlc/xgboost）
- 对于神经网络：keras（http://keras.io/）
- 数据绘图：matplotlib（http://matplotlib.org/）
- 监视进度：tqdm（https://pypi.python.org/pypi/tqdm）
Videos :
- 21 Deep Learning Videos, Tutorials & Courses on Youtube from 2016
- RL Course by David Silver - Lecture 1: Introduction to Reinforcement Learning
- 斯坦福2017季CS224n深度学习自然语言处理课程
- CS224d: Deep Learning for Natural Language Processing ( Doing )
- CS 294: Deep Reinforcement Learning, Fall 2017
- Morvan
- 李宏毅深度学习(2017)
Github Projects:
- Machine Learning Mindmap / Cheatsheet ( Only Pictures)
- https://github.com/dformoso/machine-learning-mindmap
- A Mindmap summarising Machine Learning concepts, from Data Analysis to Deep Learning.
DeepMind : Teaching Machines to Read and Comprehend
- https://github.com/thomasmesnard/DeepMind-Teaching-Machines-to-Read-and-Comprehend
- This repository contains an implementation of the two models (the Deep LSTM and the Attentive Reader) described in Teaching Machines to Read and Comprehend by Karl Moritz Hermann and al., NIPS, 2015. This repository also contains an implementation of a Deep Bidirectional LSTM.
A-Guide-to-DeepMinds-StarCraft-AI-Environment
- https://github.com/llSourcell/A-Guide-to-DeepMinds-StarCraft-AI-Environment
- This is the code for “A Guide to DeepMind’s StarCraft AI Environment” by Siraj Raval on Youtube
- Must install in venv to avoid wrong operation
- Keras Text Classification Library
- https://github.com/raghakot/keras-text
- keras-text is a one-stop text classification library implementing various state of the art models with a clean and extendable interface to implement custom architectures.
- emnlp
RL
- 【Keras+OpenAI增强学习实践：Actor-Critic模型】《Reinforcement Learning w/ Keras + OpenAI: Actor-Critic Models》by Yash Patel O网页链接 pdf:O网页链接
- 《Keras+OpenAI强化学习实践：深度Q网络》via:机器之心 O教程 | Keras+OpenAI强化学习实践：深度Q网络
- 《A Tour of Gotchas When Implementing Deep Q Networks with Keras and OpenAi Gym》by Scott Rome O网页链接
- 【层次增强学习算法】《Learning a Hierarchy | OpenAI》 O网页链接 ref:《Meta Learning Shared Hierarchies》(2017) GitHub: https ://github .com/openai/mlsh
- 《OpenAI实习生提出分层强化学习新算法》via:机器之心
- +
- Q-Learning月球着陆控制“QLearning in OpenAI Lunar Lander”
- GitHub: https:\//github.com\/FitMachineLearning/FitML/
- 强化学习入门 第一讲 MDP
- 强化学习入门及其实现代码
+ http://www.jianshu.com/p/165607eaa4f9 - Course
- 如2015年David Silver的经典课程Teaching ，
- 2017年加州大学伯克利分校Levine, Finn, Schulman的课程 CS 294 Deep Reinforcement Learning, Spring 2017
- 卡内基梅隆大学的2017 春季课程Deep RL and Control 。
- 【转载】近似动态规划与强化学习入门步骤
- Awesome Reinforcement Learning
- https://github.com/aikorea/awesome-rl
- lots of resourses
- http://www.jianshu.com/p/165607eaa4f9
- keras-rl
Other Resources:
- 模型汇总16 各类Seq2Seq模型对比及《Attention Is All You Need》中技术详解
- 模型汇总24 - 深度学习中Attention Mechanism详细介绍：原理、分类及应用
- 人工智能 Java 坦克机器人系列强化学习-IBM Robo code
- 遗传算法
蒙特卡罗算法
- https://www.zhihu.com/question/20254139
- RL
- 深度强化学习（Deep Reinforcement Learning）入门：RL base & DQN-DDPG-A3C introduction
- 深度增强学习前沿算法思想【DQN、A3C、UNREAL简介】
- Google Deepmind大神David Silver带你认识强化学习
- 深度强化学习（Deep Reinforcement Learning）入门：RL base & DQN-DDPG-A3C introduction
- NLP
- 综述 | 一文读懂自然语言处理NLP（附学习资料）
- 综述 | 一文读懂自然语言处理NLP（附学习资料）
DL
- 126 篇殿堂级深度学习论文分类整理 从入门到应用 | 干货9
- DeepLearningBook读书笔记
- 126 篇殿堂级深度学习论文分类整理 从入门到应用 | 干货9
- ML
- 良心GitHub项目：各种机器学习任务的顶级结果（论文）汇总
- 良心GitHub项目：各种机器学习任务的顶级结果（论文）汇总
- Transfor Learning
- 14 篇论文为你呈现「迁移学习」研究全貌
- 14 篇论文为你呈现「迁移学习」研究全貌
- SKLearn(工程用用的较多的模块介绍)
- Tensorflow
- （较好）转TensorFlow实现案例汇集：代码+笔记
- 数十种TensorFlow实现案例汇集：代码+笔记
- 【推荐】TensorFlow/PyTorch/Sklearn实现的五十种机器学习模型
- （较好）转TensorFlow实现案例汇集：代码+笔记
- Others
- cs231n课程笔记翻译
- 全网AI和机器学习资源大合集（研究机构、视频、博客、书籍…）
- cs231n课程笔记翻译
NIPS大会最精彩一日：AlphaZero遭受质疑；史上第一场正式辩论与LeCun激情抗辩；元学习&强化学习亮点复盘
- 自然语言顶级会议ACL 2016谷歌论文汇集
- 解决机器学习问题有通法
- 比AlphaGo Zero更强的AlphaZero来了！8小时解决一切棋类！
Corpus:
- 国家语委
- 国内可用免费语料库
- 总词汇量在7400W+，可以用于训练很多模型例如Max Entropy、CRF、HMM……，优点是这份语料在分词基础上还做了词性标注，至于准确性还有待考究。
https://github.com/liwenzhu/corpusZh
扫一扫
872
被折叠的 条评论
为什么被折叠?
到【灌水乐园】发言
