20 min listen
053 - Chasing convex bodies and other random topics with Dr. Sebastien Bubeck
053 - Chasing convex bodies and other random topics with Dr. Sebastien Bubeck
ratings:
Length:
20 minutes
Released:
Dec 5, 2018
Format:
Podcast episode
Description
Dr. Sebastien Bubeck is a mathematician and a senior researcher in the Machine Learning and Optimization group at Microsoft Research. He’s also a self-proclaimed “bandit” who claims that, despite all the buzz around AI, it’s still a science in its infancy. That’s why he’s devoted his career to advancing the mathematical foundations behind the machine learning algorithms behind AI.
Today, Dr. Bubeck explains the difficulty of the multi-armed bandit problem in the context of a parameter- and data-rich online world. He also discusses a host of topics from randomness and convex optimization to metrical task systems and log n competitiveness to the surprising connection between Gaussian kernels and what he calls some of the most beautiful objects in mathematics.
Today, Dr. Bubeck explains the difficulty of the multi-armed bandit problem in the context of a parameter- and data-rich online world. He also discusses a host of topics from randomness and convex optimization to metrical task systems and log n competitiveness to the surprising connection between Gaussian kernels and what he calls some of the most beautiful objects in mathematics.
Released:
Dec 5, 2018
Format:
Podcast episode
Titles in the series (100)
001 - Notes from the Productivity Revolution with Dr. Jaime Teevan by Microsoft Research Podcast