ml-directory
Machine learning directory
Papers
Reference papers
- Batch normalization: Batch Normalization: Accelerating Deep Network Training by Reducing Internal Covariate Shift
- Dropout regularization: Dropout: A Simple Way to Prevent Neural Networks from Overfitting
- Elastic net regularization: Regularization and variable selection via the elastic net
Books
Statistics
- Think Stats: Probability and Statistics for Programmers by Allen Downey - (with free PDF download link)
Reinforcement Learning:
- Reinforcement Learning: An Introduction by Sutton & Barto
Courses
Machine Learning:
Deep stuff:
- CS231n: Convolutional Neural Networks for Visual Recognition
- CS224d: Deep Learning for Natural Language Processing
- CS 294: Deep Reinforcement Learning
ML Heuristics
Quotes
- Keep PCA components to maintain 97-98% of variance (Andrew Ng, First ML MOOC)
- k=10 in k-fold CV (Ambroise McLachlan)
- Use 10x samples the VC-dimension for training (Learning from Data course by Yaser S. Abu-Mostafa)
- Word2vec: Skip Gram works well with small amount of data and is found to represent rare words well. On the other hand, CBOW is faster and has better representations for more frequent words. Article
Papers
Papers that account heuristics:
- A Sensitivity Analysis of (and Practitioners' Guide to) Convolutional Neural Networks for Sentence Classification
- An Empirical Exploration of Recurrent Network Architectures
- LSTM: A Search Space Odyssey
- An overview of gradient descent optimization algorithms
Pages
- CNN Tricks
- How to Train a GAN? Tips and tricks to make GANs work
- Guideline to select the hyperparameters in Deep Learning
- ML Fundamentals Checklist by Microsoft
Topics
AutoML - Automated Machine Learning:
- AutoML.org
- AutoML Workshops:
- MetaSel Workshop 2014 - Meta-Learning and Algorithm Selection @ECAI
Web sites
Tutorials
- Deep Learning tutorials directory
- Kaggle tutorials directory
- Unsupervised Feature Learning and Deep Learning
- Geoffrey Hinton's Neural Networks for Machine Learning
- General Sequence Learning using Recurrent Neural Networks
- Hacker's guide to Neural Networks
- Deep Learning tutorial @ Kaggle
- Andrew W. Moore tutorials on several aspects of statistical data mining.
CNN
RNN
- RNN tutorial in 4 parts
- The Unreasonable Effectiveness of Recurrent Neural Networks
- Understanding LSTMs
Software
- https://www.r-project.org/ [R]
- http://www.scipy.org/ [Python]
- TensorFlow
- pandas - Python Data Analysis Library
- xgboost
Recommender Systems:
- Surprise - A Python scikit for building and analyzing recommender systems