ISL Colloquium presents Taming the Devil of Gradient-based Optimization Methods with the Angel of Differential Equations

Topic: 
Taming the Devil of Gradient-based Optimization Methods with the Angel of Differential Equations
Thursday, November 1, 2018 - 4:15pm
Venue: 
Packard 101
Speaker: 
Professor Weijie Su (UPenn)
Abstract / Description: 

In this talk, we use ordinary differential equations to model, analyze, and interpret gradient-based optimization methods. In the first part of the talk, we derive a second-order ODE that is the limit of Nesterov's accelerated gradient method for non-strongly objectives (NAG-C). The continuous-time ODE is shown to allow for a better understanding of NAG-C and, as a byproduct, we obtain a family of accelerated methods with similar convergence rates. In the second part, we begin by recognizing that existing ODEs in the literature are inadequate to distinguish between two fundamentally different methods, Nesterov's accelerated gradient method for strongly convex functions (NAG-SC) and Polyak's heavy-ball method. In response, we derive high-resolution ODEs as more accurate surrogates for the three aforementioned methods. These novel ODEs can be integrated into a general framework that allows for a fine-grained analysis of the discrete optimization algorithms through translating properties of the amenable ODEs into those of their discrete counterparts. As the first application of this framework, we identify the effect of a term referred to as 'gradient correction' in NAG-SC but not in the heavy-ball method, shedding insight into why the former achieves acceleration while the latter does not. Moreover, in this high-resolution ODE framework, NAG-C is shown to boost the squared gradient norm minimization at the inverse cubic rate, which is the sharpest known rate concerning NAG-C itself. Finally, by modifying the high-resolution ODE of NAG-C, we obtain a family of new optimization methods that are shown to maintain the accelerated convergence rates as NAG-C for smooth convex functions. This is based on joint work with Stephen Boyd, Emmanuel Candes, Simon Du, Michael Jordan, and Bin Shi.

Bio:

Weijie Su is currently an Assistant Professor of Statistics in the Department of Statistics at the Wharton School, University of Pennsylvania. His research interests are in statistical machine learning, high-dimensional statistical inference, large-scale multiple testing, and privacy-preserving data analysis. Prior to joining Penn in Summer 2016, He obtained his Ph.D. in Statistics from Stanford University in 2016, under the supervision of Emmanuel Candès, and received his bachelor's degree in Mathematics from Peking University in 2011.