Machine Learning Weekly Review №3 — the source of latest credible papers, videos and projects on machine learning for scientists and engineers.

Recommended this week


  1. Graph Auto-Encoders — Tensorflow implementation of Variational Graph Auto-Encoders.
  2. Mxnet-memonger — Sublinear memory optimization for deep learning, reduce GPU memory cost to train deeper nets.
  3. Facets — Open Source Visualization Tool for Machine Learning Training Data from Google Developers.


  1. “Learning Macromanagement in StarCraft from Replays using Deep Learning” — outperforms the game’s built-in Terran bot.
  2. Self Adversarial Training for Human Pose Estimation”. A deep ConvNet model learns the structure and configuration of human body parts via adversarial training.Achieves state-of-the-art results on LSP, MPII, and LIP datasets.
  3. Enhanced Deep Residual Networks for Single Image Super-Resolution” –appears in CVPR 2017 workshop. Best paper award of the NTIRE2017 workshop, and the winners of the NTIRE2017 Challenge on Single Image Super-Resolution.
  4. The Reversible Residual Network: Backpropagation Without Storing Activations”. A variant of ResNets where each layer’s activations can be reconstructed exactly from the next layer’s, so the activation storage requirements are independent of depth.
  5. “Overcoming the curse of dimensionality: Solving high-dimensional partial differential equations using deep learning”.
  6. “The Devil is in the Decoder”. Paper from Google presents an extensive comparison of a variety of decoders for a variety of pixel-wise prediction tasks.
  7. “Deep Bilateral Learning for Real-Time Image Enhancement”. Paper from Google and MIT researchers processes high-resolution images on a smartphone in milliseconds.


  1. Curated list of R tutorials for Data Science, NLP and Machine Learning
  2. Machine Learning Crash Course by UC Berkeley


  1. What are some new and exciting areas in adversarial machine learning research? Answered by Ian Goodfellow.
  2. You and Your Research” by Richard Hamming. Transcription of the Bell Communications Research Colloquium Seminar 7 March 1986.
  3. “Learning to Learn” by Berkeley AI PhD student Chelsea Finn.
  4. A curated list of super-resolution benchmarks and resources for single image super-resolution algorithms.
  5. Using Machine Learning to Predict Value of Homes On Airbnb by Robert Chang.
  6. Neuroevolution: A different kind of deep learning. The quest to evolve neural networks through evolutionary algorithms.


  1. SciPy 2017: Scientific Computing with Python Conference. Videos of the conference talks and tutorials.
  2. Parallel Programming in R and Python. Shows you how to utilize multi-core, high-memory machines to dramatically accelerate your computations in R and Python, without any complex or time-consuming setup. By Domino Data Lab.
  3. Data Science in Practice. Course by Professor Bradley Voytek UC San Diego.

You would also like

Facebook Group: Machine Learning Review