Data Sets for the study "Non-Elitist Selection Can Improve the Performance of Irace"

Zenodo (CERN European Organization for Nuclear Research)

Abstract

This is the result of the study "Non-Elitist Selection Can Improve the Performance of Irace." Apart from the results of the three methods, i.e., irace, irace-rand, and irace-entropy, presented in the paper, the results of irace-Gower, which applies the Gower's distance to measure the diversity of configurations, are also provided in this dataset. The irace variants are tested on three scenarios: ACOTSP, ACOQAP, and SPEAR. This Data set consists of: 1. The configurations obtained by 20 runs of each irace variant. 2. The results of validation runs of the obtained configurations. Contact: if you have any questions or suggestions, please get in touch with Furong Ye.

Authors 4

1. Furong Ye Aachen [LIACS

   Leiden University

   Affiliation as printed

   Liacs, Leiden University

2. Diederick Vermetten Aachen [LIACS

   Leiden University

   Affiliation as printed

   Liacs, Leiden University

3. Carola Doerr

   Sorbonne Université

   Affiliation as printed

   Sorbonne Université

4. Thomas B ̈ack Aachen [LIACS

   Leiden University

   Affiliation as printed

   Liacs, Leiden University