I will no longer update "Deep Learning" part of this note frequently. Several reasons like:

1. Fairly speaking, these are much more than enough for a beginner. If you could take good use of them, then you should be awesome now!
2. I cannot afford expensive GPUs. Deep Learning is still somewhat like an alchemy [E.g.] (You are search for this topic on Twitter).
3. I'm more interested in the application of (Statistical) Machine Learning these days. So I would probably update some Machine Learning related resources in the future.
4. I don't want to go on a PhD to do research. I believe the knowledge I have now is enough to me. Startup-like things might be more enjoyable.
5. 人生苦短，我更喜欢刺激与惊喜 (Short as life is, I pursue thrill and surprise.)
6. As pointed out by some senior researchers, I am not smart enough to compete with those people.:man_facepalming: (Half kidding)

ML_Note

A Machine Learning Resource gathering as a collection of many extremely useful resources. Feel free to contact me at if you are interested. I would be really glad if this helps you.

This note might be confusing to you now since I fail to come up with an idea to organize it. But I will try to make it universally applicable later. If you can't wait any more to study ML, maybe I can suggest you a pathway by email.

You may want to search or contibute

Table of Contents

1. Conceptual
2. Personal blogs and official accounts
3. Courses and Tutorials
4. Articles
5. Papers
   1. GAN
   2. NLP
   3. Miscellanous
6. Webs or APIs
   1. Starting Point
   2. For Python
   3. Advanced / Miscellanous
7. Datasets
8. Projects

Conceptual

Wikipedia is always your first choice!

• Math
• Basic
• Information theory
• GAN
• NLP
• Miscellaneous

Personal blogs and official accounts

• Zhihu topics: Deep Learning
• Andrew Ng
• Xf Mao
• xf__mao
• Poll
• 邓范鑫——致力于变革未来的智能技术(个人博客)
• 开发者头条
• Chris McCormick Machine Learning
• ml4a Machine Learning for Artists
• TensorFlow 官方文档中文版
• handong1587's blog
• Daniel Holdenresearcher at Ubisoft Montreal
• 《机器学习实战》学习笔记
• CURTIS MILLER'S PERSONAL WEBSITE
• 返回主页 刘建平Pinard
• Siraj Raval
• Sebastian Ruder
• Data science blogs

Courses and Tutorials

• Machine Learning - Coursera Fundamental ML
• Machine Learning - Google Highly recommended fundamental tutorial starting from Decision Tree algorithm by Google, though sadly aborted half way
• DeepLearningZeroToAll Fundamental ML. Not suitable for first learners, but good for revision.
• fast.ai
• CMU deep learning
• Stanford CS229: Machine Learning
• Stanford CS231A: Computer Vision, From 3D Reconstruction to Recognition
• Stanford CS231n: Convolutional Neural Networks for Visual Recognition
• generative-adversarial-networks Fundamental GAN
• Nvidia Deep Learning Education
• DeepLearningTutorials written in theano
• 中文 Python 笔记 Tutorials for Python 基础与进阶, Numpy, Scipy, Matplotlib, Cython, OOP, Theano, Pandas and many other useful moudules for python
• Computer VisionNYU Fall 2012
• 动手学深度学习李沐: Learning Deep Learning using Apache MXNet (incubating) 's latest API –––– gluon
• deeplearningbook
• deeplearningbook-chinese
• DeepLearningBook读书笔记
• Oxford Deep NLP 2017 course
• Machine Learning and Data Mining - by University of California in Irvine cs273a
• Scala - I cannot understand why people love this language. If you have to use this last-choice language, check this course. If you are eager, check cheetsheet Please, stop wasting your time on this [Facepalm]. If you were asked to use it, complain to your supervisor with no hesitance.
• Udacity - too many wonderful courses!
• 斯坦福大学秋季课程《深度学习理论》STATS 385, HomePage, Videos, Videos on Bilibili
• Book Collection
• UC Berkeley CS 294: Deep Reinforcement Learning, Fall 2017
• 《迁移学习简明手册》
• 迁移学习介绍、综述文章、最新文章、代表工作及其代码、常用数据集、硕博士论文

Articles

• 机器学习常见算法分类汇总
• Vanishing Gradient Problem
• GAN入门教程
• 【深度神经网络压缩】Deep Compression （ICLR2016 Best Paper）
• Complete Guide to Parameter Tuning in XGBoost (with codes in Python)
• When does parameter-sharing in recurrent neural networks make sense?
• 如何配置一台适用于深度学习的工作站？
• 深度学习从入门到出门·机器玄学文章合集
• 机器学习经典 PRML 最新 Python 代码实现，附最全 PRML 笔记视频学习资料
• 令人拍案叫绝的Wasserstein GAN

Papers

GAN

• really-awesome-gan Collection of GAN papers
• AdversarialNetsPapers Collection of GAN papers
• short summaries of some machine learning papers
• 【Ian Goodfellow 强推】GAN 进展跟踪 10 大论文（附下载）

NLP

• R-NET: Machine Reading Comprehension with Self-matching Networks

Miscellaneous

• DeepCodingBaselines Collection of ML papers on Code Completion, Dialogue, Bug Localization, Code Clone Detection, Commit Summarization, Bug Fix and Overrun Detect
• Action Recognition using Visual Attention
• 深度学习的9篇开山之作
• deep-learning-papers
• Conference Open Review
• Arxiv Sanity Preserver - Built in spare time by @karpathy to accelerate research, Serving last 36935 papers from cs.
• State-of-the-art result for all Machine Learning Problems, or this

Webs or APIs

Starting Point

• Tensorflow - faster

• Keras - Keras is now integrated in Tensorflow: tf.keras

• Pytorch - easier to use

• Theano

• MXNet - faster

• Performance Modeling and Evaluation of Distributed Deep Learning Frameworks on GPUs

• AWS EC2 - one year free for some micro servers. Education Entry - get free USD$100 each year

• Paperspace - cheaper and faster

• Some guidance if you would like to switch: Guidance 1. If you are new comers, these guidances might not help. From my own experience, just randomly choose one (I would suggest Keras and Pytorch though). Before I touched some of them, I have no idea what these articles are talking about. Even if I went through some articles, I still have no idea which one to choose.

For Python

• Anaconda - Powerful Machine Learning Package Manager
• Brew for MacOS, Brew for Linux
• pip - built-in for python, pip install pkg_name, to specify which version of python the pkg is to install for, use pip2 to pip3
• Jupyter Notebook - elegant and powerful interactive python v3.X shell(can also support for V2.0 or R, etc.)
• Scipy
• Numpy - Powerful and efficient handler for numbers, matrices, and their operations. With CLEAR documentation.
• Pandas - Powerful and efficient handler for Data I/O and preprocessing. With CLEAR documentation.
• matplotlib - Powerful tool for ploting images.
• scikit-learn - Powerful package containg many basic algorithms in Classification, Regression, Clustering, Dimensionality reduction, Model selection and Preprocessing. With CLEAR documentation.

Advanced / Miscellanous

• grakn.ai
• algorithmia
• 哈工大语言技术平台云, 新注册的ai域名
• jieba - 结巴中文分词
• NLTK - many NLP tools
• THUMT - an open-source neural machine translation toolkit developed by Tsinghua Natural Language Processing Group
• NOUS - Construction, Querying and Reasoning with Knowledge Graphs
• CN-DBpedia - 复旦·知识图谱API
• Machine Learning for Cyber Security - Datasets, Papers, Books, Talks, Tutorials, Courses, Miscellaneous
• Generate c code
• fairseq - Facebook AI Research Sequence-to-Sequence Toolkit
• PHP-ML - Machine Learning library for PHP
• WordNet - A lexical database for English
• DeepMind
• ConvNetJS - Deep Learning in your browser, by Stanford
• betafaceapi
• Kesci
• XGBoost
• rsync - fast incremental file transfer tool
• Deep Learning
• Toronto University ML
• 大牛们的blog （人工智能与机器学习）
• lab@grapeot.me
• kaggle-past-solutions
• NIPS(Neural Information Processing Systems) paper
• analysing-the-arxiv arxiv metadata fetch
• Using Jupyter notebook securely in remote cluster
• Google A.I. Experiments
• YodaQA open source question answering system
• Text Mining Online
• Stanford.NLP.NET, 简单说明
• Neo4j - graph database
• cayley - An open-source graph database
• pyley - Python client for an open-source graph database Cayley
• AI商用搜索 - AI商用搜索——By 机器之心
• Green Screen - 自动抠图, Article
• OpenKG.CN - 中文开放知识图谱
• paddlepaddle - PArallel Distributed Deep LEarning –––– Deep Learning APIs by Baidu
• MachineLearning
• ONNX - Open Neural Network Exchange. import module for Caffe2, export a network from PyTorch to ONNX
• Sequoia 2000 FTP server home page
• Game Agent Framework, http://serpent.ai
• 哈工大科大讯飞联合实验室
• OPENML
• Baidu MDL - Mobile Deep Learning
• ML Showcase
• tushare - 财经数据接口包
• 52 个超级有用的机器学习与预测接口
• Losswise - 机器学习项目监控解决方案
• call-python-libraries - Access Python^® functionality from MATLAB^®
• tensorpack - A neural net training interface based on TensorFlow
• GloVe: Global Vectors for Word Representation - have pre-trained word vectors
• Beijing City Lab
• 知网（HowNet）知识库的简单调用指南
• OpenNE - An open source toolkit for Network Embedding
• OpenKE - An Open-Source Package for Knowledge Embeddingd
• OpenCV - Computer Vision tool, can be treated as Photoshop
• FlashText - Extract Keywords from sentence or Replace keywords in sentences. Much faster than Regex on longer sentences. Instruction By 机器之心
• TensorLayer - Deep Learning and Reinforcement Learning Library for Researcher and Engineer.
• DPM v5 - Discriminatively trained deformable part models, Instruction to run DPM v5 with Pascal VOC devkit
• Install Linuxbrew on your AWS instance
• 使 Amazon EBS 卷可用
• Deep Learning for Java - Open-Source, Distributed, Deep Learning Library for the JVM
• spaCy - Industrial-strength Natural Language Processing (NLP) with Python and Cython
• DeepVis
• 13种编程语言对应的机器学习资源大全！
• 别错过这张AI商用清单：你的生产难题可能被一个应用解决（上），（下）
• ParlAI––––A unified platform for sharing, training and evaluating dialog models across many tasks.
• Pandas on Ray Pandas 改进版
• Amazon connect a new EBS
• Compute Code Diff

Datasets

• AWS Public Datasets
• SF OpenData
• Google cluster workload traces
• Kaggle dataset
• Experimental Data for Question Classification
• UC Irvine Machine Learning Repository
• Network data
• iPinYou Real-Time Bidding Dataset for Computational Advertising Research
• Chinese Cloze-style RC Dataset: People Daily & Children's Fairy Tale
• Million Song Dataset
• Chest X-ray provided by National Institutes of Health - Clinical Center
• DataSciComp - A collection of popular Data Science Competitions
• Open Academic Graph - This data set is generated by linking two large academic graphs: Microsoft Academic Graph (MAG) and AMiner.
• 各领域公开数据集下载
• VOT2016 Dataset - Visual Object Tracking
• PASCAL VOC
• S＆P 500 - STATWORX 团队近日(2017-11-12)从 Google Finance API 中精选出了 S＆P 500 数据，该数据集包含 S＆P 500 的指数和股价信息
• ImageNet, paper
• MSCOCO -paper large-scale object detection, segmentation, and captioning dataset.
• OpenSurfaces - large database of annotated surfaces created from real-world consumer photographs.
• Street View House Numbers (SVHN)
• ML Open Data
• Stanford Network Analysis Project
• Acemap Knowledge Graph
• YTF (YouTube Faces DB), login to this server, username: wolftau, PW: wtal997 (ref)

Projects

• 8个带你快速入门的趣味机器学习项目（附数据源、教程）
• 《21个项目玩转深度学习———基于TensorFlow的实践详解》配套代码
• Keras implementations of Generative Adversarial Networks, PyTorch implementations of Generative Adversarial Networks
• Implementation of Conditional WGAN and WGAN in pytorch

⬆️⬆️⬆️TOP