This post is a summary of articles, websites and material in general about machine learning.
Articles
- RNNs
- Get an overview: The Unreasonable Effectiveness of Recurrent Neural Networks
- Understand them: Understanding LSTM Networks
- Using convolutional neural nets to detect facial keypoints tutorial
- Clever Methods of Overfitting
- Understanding the Bias-Variance Tradeoff
- An overview of gradient descent optimization algorithms
- cs231n: Convolutional Neural Networks (CNNs / ConvNets) (YouTube playlist)
- Evolution Strategies
- Warning Signs in Experimental Design and Interpretation: Not the typical ML literature, but interesting and relevant nonetheless as ML is driven by experiments.
Books
- Neural Networks and Deep Learning
- Ian Goodfellow, Yoshua Bengio, and Aaron Courville: Deep Learning
MOOCs
- Coursera: Machine Learning by Andrew Ng
- CS224d: Deep Learning for Natural Language Processing
- Machine Learning: Course by the University of Oxford
- Convolutional Neural Networks for Visual Recognition: Course by Stanford
Tools
- Caffe: Used often for Computer Vision, but more and more people jump to TensorFlow
- sklearn: Python Machine learning toolkit
- Theano: Used often for Speech Recognition
- TensorFlow: C++ and Python, supports nVidia GPU training of neural networks
- Keras.io: Extremely nice for beginners
Data
Collections
Benchmark Datasets
- MNIST: 70 000 images of \(28 \times 28\) px with labels (digits 0-9)
- HASY: 168 233 images of \(32 \times 32\) px with labels (369 classes, all of them are characters)
- HWRT: Handwritten symbols (similar to HASY, but online data)
- IRIS: 3 classes, 50 items per class, 4 features per item
- KITTI: Road vision dataset
Lists:
- metacademy.org: A lot of material when you know what to look for
- computervisiononline.com: A list of many datasets
- YACVID: Computer Vision Index To Datasets
- dmoz.org
Cheat Sheets
Lists
- Machine Learning Tutorials by Ujjwal Karn (Facebook employee)
- Awesome Random Forest: A curated list of resources regarding tree-based methods and more, including but not limited to random forest, bagging and boosting.
Miscellaneous
- Kaggle: Machine Learning Challenges
- Stack Exchange
- awesome-machine-learning: A list with MANY links to machine learning tools
- Demos:
- Neural Machine Translation: English → German, French
- write-math.com: Symbol recognition
- Tensorflow Playground: Demo for decision boundary of neural network
- lecture-demo.ira.uka.de: Rosenblatt-Perceptron, GMMs, ...
- demos.algorithmia.com/colorize-photos: Colorize a grayscale photo