Construction of symmetric orthogonal designs with deep Q-network and orthogonal complementary design

Jianfa Lai, Lin-Chen Weng, Xiaoling Peng and Kai-Tai Fang

Computational Statistics & Data Analysis, 2022, vol. 171, issue C

Abstract: Construction of orthogonal designs (ODs) has received much attention over the past decades, where previous work was originated from either mathematical theory or algorithmic search. A new algorithm is proposed to construct symmetric ODs. It is established on a well-designed framework of sequential construction, combining the deep Q-network (DQN) and orthogonal complementary design (OCD). The DQN-OCD algorithm shows its superiority by constructing various non-isomorphic ODs in an efficient manner. In particular, the constructions of symmetric ODs, including the saturated ODs L27(313), L28(227) and non-saturated ODs L18(37), L36(313) are presented, where the performance of DQN-OCD algorithm surpasses the others. Furthermore, a series of previously unknown ODs in non-isomorphic subclasses of L28(227) and L36(313) are constructed as new collections of ODs.

Keywords: Deep Q-network; Design isomorphism; Orthogonal complementary design; Orthogonal design; Sequential construction (search for similar items in EconPapers)
Date: 2022
References: View references in EconPapers View complete reference list from CitEc
Citations:

Downloads: (external link)
http://www.sciencedirect.com/science/article/pii/S0167947322000287
Full text for ScienceDirect subscribers only.

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:eee:csdana:v:171:y:2022:i:c:s0167947322000287

DOI: 10.1016/j.csda.2022.107448

Access Statistics for this article

Computational Statistics & Data Analysis is currently edited by S.P. Azen

More articles in Computational Statistics & Data Analysis from Elsevier
Bibliographic data for series maintained by Catherine Liu ().