Comparison of Low-budget Black-box Optimization Algorithms on BBOB

Zenodo (CERN European Organization for Nuclear Research)

Abstract

Data to replicate the results presented in the paper "Low-budget Black-box Optimization Algorithms in BBOB and OpenAI Gym", submitted to IEEE Transactions on Evolutionary Computation. Here, we offer the data comparing Black-Box Optimization tools for machine learning with more classical heuristics on the well-known BBOB benchmark suite from the COCO environment (24 noiseless functions with different landscape characteristics: uni/multi-modality, separability, good/weak global structure, etc.)

Authors 5

1. Elena Raponi Aachen

   Leiden University

   Affiliation as printed

   Leiden University (LIACS)

2. Carraz Nathanael Rakotonirina

   University of Antananarivo

   Affiliation as printed

   Université d'Antananarivo, Madagascar

3. Rapin Jérémy
   Affiliation as printed

   Meta AI Research

4. Carola Doerr

   Centre National de la Recherche Scientifique · Sorbonne Université · LIP6

   Affiliation as printed

   Sorbonne Université (LIP6, CNRS)

5. Olivier Teytaud
   Affiliation as printed

   Meta AI Research