The Role of Memory in Stochastic Optimization
METADATA ONLY
Loading...
Author / Producer
Date
2020
Publication Type
Conference Paper
ETH Bibliography
yes
Citations
Altmetric
METADATA ONLY
Data
Rights / License
Abstract
The choice of how to retain information about past gradients dramatically affects the convergence properties of state-of-the-art stochastic optimization methods, such as Heavy-ball, Nesterov’s momentum, RMSprop and Adam. Building on this observation, we use stochastic differential equations (SDEs) to explicitly study the role of memory in gradient-based algorithms. We first derive a general continuous-time model that can incorporate arbitrary types of memory, for both deterministic and stochastic settings. We provide convergence guarantees for this SDE for weakly-quasi-convex and quadratically growing functions. We then demonstrate how to discretize this SDE to get a flexible discrete-time algorithm that can implement a board spectrum of memories ranging from short- to long-term. Not only does this algorithm increase the degrees of freedom in algorithmic choice for practitioners but it also comes with better stability properties than classical momentum in the convex stochastic setting. In particular, no iterate averaging is needed for convergence. Interestingly, our analysis also provides a novel interpretation of Nesterov’s momentum as stable gradient amplification and highlights a possible reason for its unstable behavior in the (convex) stochastic setting. Furthermore, we discuss the use of long term memory for second-moment estimation in adaptive methods, such as Adam and RMSprop. Finally, we provide an extensive experimental study of the effect of different types of memory in both convex and nonconvex settings.
Permanent link
Publication status
published
Book title
Proceedings of The 35th Uncertainty in Artificial Intelligence Conference
Journal / series
Volume
115
Pages / Article No.
356 - 366
Publisher
PMLR
Event
35th Conference on Uncertainty in Artificial Intelligence (UAI 2019)
Edition / version
Methods
Software
Geographic location
Date collected
Date created
Subject
Organisational unit
09462 - Hofmann, Thomas / Hofmann, Thomas