Machine Learning Weekly Review №2

Source of latest credible papers, videos and projects on machine learning for scientists and engineers.

Recommended this week


  1. Neural Network Turorial from basic to hard. Available in both TensorFlow or PyTorch.
  2. Neural Machine Translation (seq2seq) Tutorial by TensorFlow Team.


  1. The Confluence of Geometry and Learning. Researchers from UC Berkeley infer 3D from 2D.
  2. UW’s lip-syncing Obama demonstrates new technique to turn audio clips into realistic video.
  3. Under the Hood of a Self-Driving Taxi. A look at compute and other core self-driving car systems by Oliver Cameron.
  4. Distributions of words across narrative time in 27,266 novels. By Stanford’s Literary Lab.


  1. Dual Supervised Learning” paper exploits correlation between dual tasks like English-to-French translation vs. French-to-English translation, speech recognition vs. text to speech etc.
  2. “Text Summarization Techniques: A Brief Survey” by Georgia Tech.
  3. “End-to-End Learning of Semantic Robotic Grasping” by GoogleBrain Team. Inspired by Two-streams hypothesis of the neural processing of vision as well as hearing.
  4. “Customer Lifetime Value Prediction Using Embeddings” deployed at
  5. “DeepStory: Video Story QA by Deep Embedded Memory Networks” introduces Deep Embedded Memory Networks (DEMN) that achieves state-of-the-art results on the MovieQA benchmark. IJCAI 2017 accepted paper.


  1. TFStage — Tensorflow Project Scaffolding. A fast and canonical project setup for models.
  2. Chainer-GAN-lib — python implementation of recent state-of-art GANs: DCGAN, BEGAN, DRAGAN etc.
  3. Scikit-plot — an intuitive library to add plotting functionality to scikit-learn objects.
  4. JupyShare — gives you a public URL for your local Jupyter notebook for easier collaboration. By Bianca.
  5. UMAP — Uniform Manifold Approximation & Projection. Better alternative to t-SNE.
  6. SentEval — a python tool by Facebook for evaluating the quality of sentence embeddings.


  1. Practical Deep Learning For Coders Course taught by Jeremy Howard (Kaggle’s #1 competitor 2 years running)
  2. Computational Statistics Tutorial from SciPy 2017 Conference. Given by Allen Downey, Professor at Olin College, author of Think Python.
  3. Differential Equations and Linear Algebra. Lectures by MIT Prof. Gilbert Strang.

You would also like

Facebook Group: Machine Learning Review