Utilizing Mixed Training and Multi-Head Attention to Address Data Shift in AI-Based Electromagnetic...
Utilizing Mixed Training and Multi-Head Attention to Address Data Shift in AI-Based Electromagnetic Solvers for Nano-Structured Metamaterials
About this item
Full title
Author / Creator
Zeng, Zhenjia , Wang, Lei , Wu, Yiran , Hu, Zhipeng , Evans, Julian , Zhu, Xinhua , Ye, Gaoao and He, Sailing
Publisher
Basel: MDPI AG
Journal title
Language
English
Formats
Publication information
Publisher
Basel: MDPI AG
Subjects
More information
Scope and Contents
Contents
When designing nano-structured metamaterials with an iterative optimization method, a fast deep learning solver is desirable to replace a time-consuming numerical solver, and the related issue of data shift is a subtle yet easily overlooked challenge. In this work, we explore the data shift challenge in an AI-based electromagnetic solver and presen...
Alternative Titles
Full title
Utilizing Mixed Training and Multi-Head Attention to Address Data Shift in AI-Based Electromagnetic Solvers for Nano-Structured Metamaterials
Authors, Artists and Contributors
Author / Creator
Identifiers
Primary Identifiers
Record Identifier
TN_cdi_doaj_primary_oai_doaj_org_article_9ebd92668c934e799db412073e66e249
Permalink
https://devfeature-collection.sl.nsw.gov.au/record/TN_cdi_doaj_primary_oai_doaj_org_article_9ebd92668c934e799db412073e66e249
Other Identifiers
ISSN
2079-4991
E-ISSN
2079-4991
DOI
10.3390/nano13202778
