Comparing Machine Learning Approaches for Context-Aware Composition
2011 (English)In: Software Composition / [ed] Sven Apel, Ethan Jackson, Springer, 2011, 18-33 p.Conference paper (Refereed)
Context-Aware Composition allows to automatically select optimal variants of algorithms, data-structures, and schedules at runtime using generalized dynamic Dispatch Tables. These tables grow exponentially with the number of significant context attributes. To make Context-Aware Composition scale, we suggest four alternative implementations to Dispatch Tables, all well-known in the field of machine learning: Decision Trees, Decision Diagrams, Naive Bayes and Support Vector Machines classifiers. We assess their decision overhead and memory consumption theoretically and practically in a number of experiments on different hardware platforms. Decision Diagrams turn out to be more compact compared to Dispatch Tables, almost as accurate, and faster in decision making. Using Decision Diagrams in Context-Aware Composition leads to a better scalability, i.e., Context-Aware Composition can be applied at more program points and regard more context attributes than before.
Place, publisher, year, edition, pages
Springer, 2011. 18-33 p.
Lecture Notes in Computer Science, ISSN 0302-9743 (print), 1611-3349 (online) ; 6708
Machine learning, context-aware composition, performance-aware components, software composition, dispatch table compression, autotuning, automated performance tuning
IdentifiersURN: urn:nbn:se:liu:diva-93373DOI: 10.1007/978-3-642-22045-6_2ISBN: 978-3-642-22044-9 (print)ISBN: 978-3-642-22045-6 (online)OAI: oai:DiVA.org:liu-93373DiVA: diva2:624366
10th International Conference on Software Composition, SC 2011; Zurich; Switzerland