Title
Lazy Hypervolume Subset Selection Algorithm with Contributions Update
Authors
[ 1 ] Instytut Informatyki, Wydział Informatyki i Telekomunikacji, Politechnika Poznańska | [ P ] employee
Scientific discipline (Law 2.0)
Year of publication
2025
Chapter type
chapter in monograph / paper
Publication language
english
Keywords
EN
- multiobjective optimization
- hypervolume
- hypervolume subset selection
Abstract
EN Hypervolume subset selection (HSS) is an important problem in evolutionary multiobjective optimization (EMO). It may be used to guide EMO algorithms, to bound Pareto archives, or to select a reduced set of the most representative solutions of a multiobjective problem for further analysis by the decision maker. Lazy greedy incremental and decremental hypervolume subset selection algorithms are currently the fastest approximate methods for more than 4 objectives and moderate sizes of the candidate sets. In this paper, we show that the efficiency of these lazy algorithms could be further improved by adaptive update of previously calculated hypervolume contributions instead of always recalculating the contributions from scratch. We show also that the Improved Quick Hypervolume algorithm is well-suited for such a context because its recursion tree does not depend on dominated points which often appear when hypervolume contributions are updated.
Date of online publication
11.08.2025
Pages (from - to)
223 - 226
Ministry points / chapter
20
Ministry points / conference (CORE)
140
System created by Poznań University of Technology
and Poznan Supercomputing and Networking Center
Log in through eKonto to add to SIS