Places I frequent
- Andrej Karpathy's blog. The probably most famous Deep Learning worker's academic blog about ideas, works and developments. I simply admire people who do solid work and share about it. Highlights:
- The Unreasonable Effectiveness of Recurrent Neural Networks
- His talk at London DL Meetup 2015: Recurrent Neural Netowrk Escapades
- The Torch Blog. As the name suggests.
- Christopher Olah's blog about understanding LSTM and more.
- Yarin Gal's blog about linking neural networks to probability models and more.
- Thinking machines. Blog by Luke Hewitt from MIT.
- The morning paper. A paper suggestion to read every morning.
MOOCs I recently signed up (or secretly follow)
- Stanford CS231n: Convolutional Neural Networks for Visual Recognition
- Stanford CS224d: Deep Learning for Natural Language Processing
- Deep Learning Summear School 2015 in Montreal
- Videos available at here
- Bayesian Learning and Gaussian Process
- Deep Learning offered by Google, using TensorFlow
- MLSS Iceland 2014. I am particularly interested in the submodularity talks:
Useful (or not but interesting) Links
- "The Believers", a lengthy but fun-to-read review of the story behind Hinton and his Deep Learning nets.
- An interview of Ilya Sutskever on topics like unsupervised learning.
- A "How the Brain Works" talk given by Hinton, June 2015
- On speeding up RNN, by Baidu Research.