Benchmarking and analyzing iterative optimization heuristics with IOHprofiler

Genetic and Evolutionary Computation Conference Companion (GECCO Companion), pp. 1043–1054

Abstract

IOHprofiler -PhilosophyUser to choose üWhich benchmark problems and algorithms to use üEasy to add new benchmark functions and algorithms üWhich performance measures to use üWide range of performance statistics: üFixed-target runtimes (+distributions) üFixed-budget results (+distributions) üProbabilities of success, ECDF curves, etc. üWhich statistical procedures to use [in progress] üKolmogorov-Smirnov, Friedman Test, Deep Statistical Comparison 1 , etc.üHow to aggregate results üOver target values, functions, and dimensions.. üHighly interactive

Authors 4

1. Hao Wang

   Sorbonne Université

   Affiliation as printed

   Sorbonne Université, Paris, France

2. Carola Doerr

   Sorbonne Université

   Affiliation as printed

   Sorbonne Université, Paris, France

3. Ofer M. Shir

   Tel Hai Academic College

   Affiliation as printed

   Tel-Hai College and Migal Institute, Upper Galilee, Israel

4. Thomas Bäck Aachen

   Leiden University

   Affiliation as printed

   Leiden University, The Netherlands