On Gradient-Based Optimization: Accelerated, Stochastic and Nonconvex
- Michael Jordan | UC Berkeley
Many new theoretical challenges have arisen in the area of gradient-based optimization for large-scale statistical data analysis, driven by the needs of applications and the opportunities provided by new hardware and software platforms. I discuss several recent, related results in this area: (1) a new framework for understanding Nesterov acceleration, obtained by taking a continuous-time, Lagrangian/Hamiltonian/symplectic perspective, (2) a discussion of how to escape saddle points efficiently in nonconvex optimization, and (3) the acceleration of Langevin diffusion.
Speaker Details
Michael I. Jordan is the Pehong Chen Distinguished Professor in the Department of Statistics and the Department of Electrical Engineering and Computer Science at the University of California, Berkeley. His research interests bridge the computational, statistical, cognitive and biological sciences. Prof. Jordan is a member of the National Academy of Sciences, the National Academy of Engineering and the American Academy of Arts and Sciences. He has been named a Neyman Lecturer and a Medallion Lecturer by the Institute of Mathematical Statistics. He received the IJCAI Research
Excellence Award in 2016 and the David E. Rumelhart Prize in 2015.
Series: MSR AI Distinguished Lectures and Fireside Chats
-
AI and Gaming Research Summit 2021 - Fireside chat with Peter Lee and Kareem Choudhry
- Peter Lee,
- Kareem Choudhry
-
Frontiers in Machine Learning: Fireside Chat
- Christopher Bishop,
- Peter Lee
-
Learning over sets, subgraphs, and streams: How to accurately incorporate graph context
- Jennifer Neville,
- Paul Bennett,
- Debadeepta Dey
-
Fireside Chat with Aaron Courville
- Aaron Courville,
- Susan Dumais
-
Fireside Chat with Maarten de Rijke
- Maarten de Rijke,
- Susan Dumais
-
First-person Perception and Interaction
- Eric Horvitz,
- Kristen Grauman
-
Fireside Chat with Anca Dragan
- Anca Dragan and Eric Horvitz
-
Conversations Based on Search Engine Result Pages
- Maarten de Rijke
-
Fireside Chat with Michael Kearns
- Michael Kearns,
- Eric Horvitz
-
The Ethical Algorithm
- Michael Kearns
-
Fireside Chat with Stefanie Jegelka
- Alekh Agarwal,
- Stefanie Jegelka
-
Fireside Chat with Peter Stone
- Peter Stone,
- Eric Horvitz
-
-
Fireside Chat with Christopher Manning
- Susan Dumais,
- Christopher Manning
-
Building Neural Network Models That Can Reason
- Christopher Manning
-
Fireside Chat with David Blei
- David Blei,
- Susan Dumais
-
The Blessings of Multiple Causes
- David Blei
-
As We May Program
- Peter Norvig
-
Fireside Chat with Peter Norvig
- Eric Horvitz,
- Peter Norvig
-
-
Fireside Chat with Manuel Blum
- Eric Horvitz,
- Manuel Blum
-
-
Fireside Chat with Dario Amodei
- Dario Amodei,
- Eric Horvitz
-
Fireside Chat with Tuomas Sandholm
- Eric Horvitz,
- Tuomas Sandholm
-
Super-Human AI for Strategic Reasoning
- Tuomas Sandholm