- This event has passed.
Emtiyaz Khan (Riken, Tokyo) – “Fast yet Simple Natural-Gradient Variational Inference”
June 27, 2:00 pm - 3:15 pm
The Statistical Seminar: Every Monday at 2:00 pm.
Time: 2:00 pm – 3:15 pm
Date: 27th of June 2018 exceptionally Wednesday
Place: Room 3001.
Emtiyaz Khan (Riken, Tokyo) – “Fast yet Simple Natural-Gradient Variational Inference in Complex Models“
Abstract: Approximate Bayesian inference is promising in improving generalization and reliability of deep learning, but is computationally challenging. Modern variational-inference (VI) methods circumvent the challenge by formulating Bayesian inference as an optimization problem and then solving it using gradient-based methods. In this talk, I will argue in favor of natural-gradient approaches which can improve convergence of VI by exploiting the information geometry of the solutions. I will discuss a fast yet simple natural-gradient method obtained by using a duality associated with exponential-family distributions. I will summarize some of our recent results on Bayesian deep learning, where natural-gradient methods lead to an approach which gives simpler updates than existing VI methods while performing comparably to them.
Cristina BUTUCEA, Alexandre TSYBAKOV, Eric MOULINES, Mathieu ROSENBAUM