Compiling Machine Learning Algorithms with SystemML Extended Abstract

Analytics on big data may range from passenger vol-ume prediction in transportation to customer satisfac-tion in automotive diagnostic systems, and from cor-relation analysis in social media data to log analysis in manufacturing. Expressing and running these an-alytics for varying data characteristics and at scale is challenging. To address these challenges, SystemML implements a declarative, high-level language using an R-like syntax extended with machine learning specific constructs, that is compiled to a MapReduce runtime [1]. The language is rich enough to express a wide class of statistical, predictive modeling and machine learning algorithms (Figure 1). We chose algorithms that are robust and scale to large, and potentially sparse data with large numbers of features