COCO: The Experimental Procedure

Nikolaus Hansen (1) , Tea Tušar (2) , Olaf Mersmann (3) , Anne Auger (1) , Dimo Brockhoff (2)

1 TAO - Machine Learning and Optimisation
2 DOLPHIN - Parallel Cooperative Multi-criteria Optimization
3 TU - Technische Universität Dortmund [Dortmund]

We present a budget-free experimental setup and procedure for benchmarking numerical optimization algorithms in a black-box scenario. This procedure can be applied with the COCO benchmarking platform. We describe initialization of and input to the algorithm and touch upon the relevance of termination and restarts.

Mots clés: Black-box; Optimization; Numerical optimization; Benchmarking

Domaines: Réseau de neurones [cs.NE]; Analyse numérique [cs.NA]; Performance et fiabilité [cs.PF]; Logiciel mathématique [cs.MS]; Apprentissage [cs.LG]

Nikolaus Hansen, Tea Tušar, Olaf Mersmann, Anne Auger, Dimo Brockhoff. COCO: The Experimental Procedure. 2016. ⟨hal-01294167v2⟩