Deep Parameter Confjguration Lars Kotthofg University of British - PowerPoint PPT Presentation

Deep Parameter Confjguration Lars Kotthofg University of British Columbia larsko@cs.ubc.ca Dagstuhl Seminar Algorithm Selection and Confjguration 1 / 4

Big Picture Algorithm Deep Parameters Exposed Parameters Algorithm Automatic Confjgurator Best Confjguration report performance propose confjguration 2 / 4

Programming by Optimization 3 / 4 Levels of PbO: Level 4: Make no design choice prematurely that cannot be justified compellingly. Level 3: Strive to provide design choices and alternatives. Level 2: Keep and expose design choices considered during software development. Level 1: Expose design choices hardwired into existing code (magic constants, hidden parameters, abandoned design alternatives). Level 0: Optimise settings of parameters exposed by existing software. Hoos & Hutter: Programming by Optimization 12

Components Mark Harman, Yue Jia, and Jens Krinke. “Deep Parameter Optimisation.” In Proceedings of the 2015 Annual Conference on Genetic and Evolutionary Computation, 1375–82. GECCO ’15. New York, NY, USA: ACM, 2015. Kevin Leyton-Brown. “Sequential Model-Based Optimization for General Algorithm Confjguration.” In LION 5, 507–23, 2011. 4 / 4 ▷ Automatic parameter extraction, Wu, Fan, Westley Weimer, ▷ Automatic confjguration, Hutter, Frank, Holger H. Hoos, and