- This event has passed.
Virtual LANS Seminar
November 11, 2020 @ 10:30 - 11:30 CST
Seminar Title: When Do We Stop SGD?
Speaker: Vivak Patel, Assistant Professor, Department of Statistics, University of Wisconsin — Madison
Date/Time: November 11, 2020 / 10:30 AM – 11:30 AM
Location: See meeting URL on the cels-seminars website (requires Argonne login)
Description: In this talk, we will motivate the need for stopping criteria for stochastic gradient descent; subsequently, we highlight and address the four main challenges for the rigorous development of such criteria: strong convergence, detectability, controlling false positive rates, and controlling false negative rates. We note that the result on strong convergence is currently the most general result of its kind for SGD on nonconvex functions, and its proof employs techniques that may be of independent interest to those who study stochastic algorithms.
Please note that the meeting URL for this event can be seen on the cels-seminars website, which requires an Argonne login.