Selected topics in deep learning

Selected topics in deep learning

Dr. G. Montufar

Max-Planck Institute for Mathematics in the Sciences

Abstract

This is a short course concerning some selected topics in deep learning. The course has three blocks. Each block will take about 2 hours. The following are the topics that will be covered in this course.

·Representational power of feedforward neural networks and probabilistic models;

·Reinforcement learning (Structure of the optimization landscape, how to choose an approximation neural network);

·Unsupervised training;

·Dimension of related algebraic varieties.

Place: Lecture Room M842, 8th floor, Beijing Institute for Scientific and Engineering Computing (BISEC), BJUT


Time: 12/Jul 14:00-16:00, 13/Jul 14:00-16:00, 14/Jul 14:00-16:00

Dr. G. Montufar received his Ph.D in Mathematics from Leipzig Univerisity in 2012. He is currently a post-doc researcher at Max-Planck Institute for Mathematics in the Sciences, and will become an Assistant Professor at UCLA soon. His research interests are very broad, including deep learning, design of learning systems, data compression, graphical models, information geometry and so on.