Machine Learning for Systems

Abstract

The six papers in this special section focus on machine learning for computer systems. Specialized computer systems have driven the performance and capability of deep learning over the past decade.1 However, as machine learning models and systems improve, there is a growing opportunity to also use these models to improve how we design, architect, optimize, and automate computer systems and software. This is a challenging area, both from a learning and a systems perspective. Systems often impose tight size, latency, or reliability constraints on learning mechanisms that do not arise in other applications of machine learning, such as computer vision or natural language processing. From a learning perspective, systems is a challenging application, where input features are often large and sparse, action spaces are gigantic, and generalization is a key attribute.