Asynchronous Parallel Stochastic Gradient Descent - A Numeric Core for Scalable Distributed Machine Learning Algorithms

Abstract

The implementation of a vast majority of machine learning (ML) algorithms boils down to solving a numerical optimization problem. In this context, Stochastic Gradient Descent (SGD) methods have long proven to provide good results, both in terms of convergence and accuracy. Recently, several parallelization approaches have been proposed in order to scale SGD to solve very large ML problems. At their core, most of these approaches are following a map-reduce scheme. This paper presents a novel parallel updating algorithm for SGD, which utilizes the asynchronous single-sided communication paradigm. Compared to existing methods, Asynchronous Parallel Stochastic Gradient Descent (ASGD) provides faster (or at least equal) convergence, close to linear scaling and stable accuracy.

Cited by in corpus (5)

• Demystifying Parallel and Distributed Deep Learning: An In-Depth Concurrency Analysis
• Communication trade-offs for synchronized distributed SGD with large step size
• Gradient Sparification for Asynchronous Distributed Training
• How Much Are You Willing to Share? A "Poker-Styled" Selective Privacy Preserving Framework for Recommender Systems
• Balancing the Communication Load of Asynchronously Parallelized Machine Learning Algorithms