Deep Learning: The Good, the Bad and the Ugly

Łukasz Kaiser (Google Brain).

Course Summary | About the lecturer | Location and schedule

Materials | Assignment

Course summary:

Neural networks made a spectacular comeback recently under the name of deep learning. We start with the basics of neural networks and dive in depth into the main factors credited for their success: faster hardware, more data and a lot of engineering. After examining those, we present some more surprising results, such as neural networks doing program synthesis with little data. We introduce a systematic classification of deep learning architectures by their expressive power and relate it to Kolmogorov complexity. This turn out to have practical applications: during the course we use an expressive network to create a high-quality translation program and help you solve a task of your choice. We also highlight the open-source side of deep learning and give guidance on how to join the community.

About the lecturer:

Łukasz joined Google in 2013 and is currently a Research Scientist in the Google Brain Team in Mountain View, where he works on fundamental aspects of deep learning and natural language processing. He has co-designed state-of-the-art neural models for machine translation, parsing and other algorithmic and generative tasks and co-authored the TensorFlow system and the Tensor2Tensor library. Before joining Google, Łukasz was a tenured researcher at University Paris Diderot and worked on logic and automata theory. He received his PhD from RWTH Aachen University in 2008 and his MSc from the University of Wroclaw, Poland.

Location and schedule:

Thursday, June 28
14:15 - 15:45 lecture
15:45 - 16:00coffee and cake
16:00 - 17:00 lecture
Friday, June 29
10:15 - 11:45 lecture
11:45 - 12:45 lunch break
12:45 - 14:00 lecture
14:00 - 14:15 coffee and cake
14:15 - 15:45 lecture
15:45 - 16:00 coffee and cake
16:00 - 17:00 lecture
Saturday, June 30
10:00 - 10:15 coffee and cake
10:15 - 11:30 lecture
11:30 - 11:45 coffee and cake
11:45 - 13:00 lecture


Assignment: here