Name
Abstract | ||
|---|---|---|
We present a dynamical system framework for understanding Nesterov's accelerated gradient method. In contrast to earlier work, our derivation does not rely on a vanishing step size argument. We show that Nesterov acceleration arises from discretizing an ordinary differential equation with a semi-implicit Euler integration scheme. We analyze both the underlying differential equation as well as the discretization to obtain insights into the phenomenon of acceleration. The analysis suggests that a curvature-dependent damping term lies at the heart of the phenomenon. We further establish connections between the discretized and the continuous-time dynamics. |
Year | Venue | Field |
|---|---|---|
2019 | International Conference on Machine Learning | Gradient method,Applied mathematics,Discretization,Differential equation,Mathematical optimization,Ordinary differential equation,Euler method,Dynamical systems theory,Acceleration,Mathematics,Dynamical system |
DocType | Volume | Citations |
Journal | abs/1905.07436 | 1 |
PageRank | References | Authors |
0.35 | 0 | 2 |
Name | Order | Citations | PageRank |
|---|---|---|---|
| Michael Muehlebach | 1 | 28 | 5.61 |
| Michael I. Jordan | 2 | 31220 | 3640.80 |