Algorithm Selection Library

The problem of algorithm selection is attracting increasing attention from researchers and practitioners from a variety of different backgrounds. After years of fruitful applications in a number of domains, a lot of data has been generated, but the community lacks a standard format or repository for this data. This situation makes it difficult to effectively share and compare different approaches, in contrast with other, more established fields. It also unnecessarily hinders new researchers seeking to begin work in this area.

We present a standardized format for representing algorithm selection scenarios and an algorithm selection library, ASlib, that contains a growing number of data sets from the literature. Our format has been designed to be able to express a wide variety of different scenarios.

The ASlib consists of algorithm selection scenarios from a many different domains, such as SAT, CSP, ASP and more. For the formal definition of the format, please see aslib.net.

Bernd Bischl -- Ludwig-Maximilians-University Munich, Germany bernd.bischl@stat.uni-muenchen.de

Lars Kotthoff -- University of British Columbia, Canada larsko@cs.ubc.ca

Marius Lindauer -- University of Freiburg, Germany lindauer@informatik.uni-freiburg.de

• Roberto Amadini - University of Bologna
• Bernd Bischl - Ludwig-Maximilians-University Munich
• Barry Hurley - Insight Centre for Data Analytics, Cork, Ireland
• Pascal Kerschke - University of Münster
• Lars Kotthoff - University of British Columbia
• Marius Lindauer - University of Freiburg
• Yuri Malitsky - IBM Thomas J. Watson Research Center
• Alexandre Frechette - University of British Columbia
• Holger Hoos - University of British Columbia
• Frank Hutter - University of Freiburg
• Kevin Leyton-Brown - University of British Columbia
• Jacopo Mauro - University of Bologna
• Kevin Tierney - University of Paderborn
• Joaquin Vanschoren - Eindhoven University of Technology

GPLv3