There was an author coming up a lot in the papers I read recently; and I think more importantly, he is one of the few (if not the only one) to have worked with both Hinton and Tenenbaum (two cognitive science / machine learning researchers I love. Yes, my love stories are weird).
I enjoyed those lectures a lot. He synthesises a lot of ideas together that are often found separately.
Kick-ass papers he authored/co-authored:
- Skip-Thought Vectors
- Dropout: A simple way to prevent neural networks from overfitting
- Human-level concept learning through probabilistic program induction
- Semantic Hashing
- Deep Boltzmann Machines
So much good stuff.