How to escape saddle points efficiently
Many new theoretical challenges have arisen in the area of gradient-based optimization for large-scale data analysis, driven by the needs of applications and the opportunities provided by new hardware and software platforms. Michael Jordan shares recent research on the avoidance of saddle points in high-dimensional nonconvex optimization.
| Talk Title | How to escape saddle points efficiently |
| Speakers | Michael Jordan (UC Berkeley) |
| Conference | Artificial Intelligence Conference |
| Conf Tag | Put AI to Work |
| Location | San Francisco, California |
| Date | September 18-20, 2017 |
| URL | Talk Page |
| Slides | |
| Video | Talk Video |
Many new theoretical challenges have arisen in the area of gradient-based optimization for large-scale data analysis, driven by the needs of applications and the opportunities provided by new hardware and software platforms. Drawing on work undertaken with Chi Jin, Rong Ge, Praneeth Netrapalli, and Sham Kakade, Michael Jordan shares recent research on the avoidance of saddle points in high-dimensional nonconvex optimization.
