A Novel Approach to Enhance DIRECT -Type Algorithms for Hyper-Rectangle Identification

Nazih-Eddine Belkacem, Lakhdar Chiter () and Mohammed Louaked
Additional contact information
Nazih-Eddine Belkacem: Department of Mathematics, Faculty of Sciences, Ferhat-Abbas University of Sétif 1, Sétif 19000, Algeria
Lakhdar Chiter: Department of Mathematics, Faculty of Sciences, Ferhat-Abbas University of Sétif 1, Sétif 19000, Algeria
Mohammed Louaked: Laboratoire de Mathématiques Nicolas Oresme, Université de Caen, Campus II, Boulevard Maréchal Juin, B.P. 5186, 14032 Caen, France

Mathematics, 2024, vol. 12, issue 2, 1-24

Abstract: This paper introduces novel enhancements to the most recent versions of DIRECT-type algorithms, especially when dealing with solutions located at the hyper-rectangle vertices. The BIRECT algorithm encounters difficulties in efficiently sampling points at the boundaries of the feasible region, leading to potential slowdowns in convergence. This issue is particularly pronounced when the optimal solution resides near the boundary. Our research explores diverse approaches, with a primary focus on incorporating a grouping strategy for hyper-rectangles of similar sizes. This categorization into different classes, constrained by a predefined threshold, aims to enhance computational efficiency, particularly involving a substantial number of hyper-rectangles of varying sizes. To further improve the new algorithm’s efficiency, we implemented a mechanism to prevent oversampling and mitigate redundancy in sampling at shared vertices within descendant sub-regions. Comparisons with several DIRECT-type algorithms highlight the promising nature of the proposed algorithms as a global optimization solution. Statistical analyses, including Friedman and Wilcoxon tests, demonstrated the effectiveness of the improvements introduced in this new algorithm.

Keywords: global optimization; derivative-free global optimization; diagonal partitioning scheme; DIRECT-type algorithms; potentially optimal hyper-rectangles (search for similar items in EconPapers)
JEL-codes: C (search for similar items in EconPapers)
Date: 2024
References: View references in EconPapers View complete reference list from CitEc
Citations:

Downloads: (external link)
https://www.mdpi.com/2227-7390/12/2/283/pdf (application/pdf)
https://www.mdpi.com/2227-7390/12/2/283/ (text/html)

Related works:
This item may be available elsewhere in EconPapers: Search for items with the same title.

Export reference: BibTeX RIS (EndNote, ProCite, RefMan) HTML/Text

Persistent link: https://EconPapers.repec.org/RePEc:gam:jmathe:v:12:y:2024:i:2:p:283-:d:1319616

Access Statistics for this article

Mathematics is currently edited by Ms. Emma He

More articles in Mathematics from MDPI
Bibliographic data for series maintained by MDPI Indexing Manager ().