deeplearning.ai's Heroes of Deep Learning: Ruslan Salakhutdinov

[@duykhanh7746]

A hero of DL making videos about "Heroes of DL"

[@vtrandal]

Main points. 1. GPUs ==> straight back propagation with much shorter training times. 2. Unsupervised learning is where advancements are needed. 3. Trends (in 2016): Deep Reinforcement Learning, Reasoning and Natural Language Understanding (Processing?), Learning from fewer examples (without lots of labeled examples).

[@berkakipek8991]

So motivational video. thanks for share.

[@argynkuketayev4166]

At 1:50 which paper exactly they're talking about? Is it "Reducing the Dimensionality of Data with Neural Networks" in Science mag?

[@sawanrai6424]

dl.acm.org/citation.cfm?id=1273596

[@evankim4096]

I wish Ruslan got asked more about deep reinforcement learning!

[@FidelaBohemia]

If you want to discuss this topic, you can read this article ... it's very interesting www.linkedin.com/pulse/deep-learning-ai-bubble-bursting-samer-l-hijazi/

[@dotnet364]

wheres ilya? he IS the hero of DL