Abstract | ||
---|---|---|
The contribution of this paper is fourfold. First, we present an updated implementation of the Improved Quick Hypervolume algorithm which is several times faster than the original implementation and according to the presented computational experiment it is at least competitive to other state-of-the-art codes for hypervolume computation. Second, we present a Greedy Decremental Lazy Quick Hypervolume Subset Selection algorithm. Third, we propose a modified Quick Hypervolume Extreme Contributor/Contribution algorithm using bounds from previous iterations of a greedy hypervolume subset selection algorithm. According to our experiments these two methods perform the best for greedy decremental hypervolume subset selection. Finally, we systematically compare performance of the fastest algorithms for greedy incremental and decremental hypervolume subset selection using two criteria: CPU time and the quality of the selected subset. |
Year | DOI | Venue |
---|---|---|
2022 | 10.1007/978-3-031-14721-0_12 | PARALLEL PROBLEM SOLVING FROM NATURE - PPSN XVII, PPSN 2022, PT II |
Keywords | DocType | Volume |
Multiobjective optimization, Hypervolume, Greedy algorithms | Conference | 13399 |
ISSN | Citations | PageRank |
0302-9743 | 0 | 0.34 |
References | Authors | |
0 | 2 |
Name | Order | Citations | PageRank |
---|---|---|---|
A. Jaszkiewicz | 1 | 660 | 50.68 |
Zielniewicz Piotr | 2 | 0 | 0.34 |