Peer-Reviewed Publication
Sci Rep2026June 2, 2026Journal Article
Modeling the interpretable geometric-performance relationship of metamaterials on small datasets using Kolmogorov-Arnold operator informed network.
Shengyu Ni1,2, Xingyi Feng1,2, Yuxin Long1,2, Xiangrong Xu1,2, Yifei Qian1,2, Hong Zhang3,4, Xiufang Gong5,6, Junyan Zhu7, Yongjie Liu1,2, Qingyuan Wang1,2
1Failure Mechanics and Engineering Disaster Prevention Key Laboratory of Sichuan Province, College of Architecture and Environment, Sichuan University, Chengdu, 610065, China.
2Key Laboratory of Deep Underground Science and Engineering, Ministry of Education, Sichuan University, Chengdu, 610065, China.
3Failure Mechanics and Engineering Disaster Prevention Key Laboratory of Sichuan Province, College of Architecture and Environment, Sichuan University, Chengdu, 610065, China. zzhanghong@scu.edu.cn.
4Key Laboratory of Deep Underground Science and Engineering, Ministry of Education, Sichuan University, Chengdu, 610065, China. zzhanghong@scu.edu.cn.
5State Key Laboratory of Clean and Efficient Turbomachinery Power Equipment, Deyang, 618000, China.
6Dongfang Electric Corporation Dongfang Turbine Co., LTD, Deyang, 618000, China.
7TaiHang Laboratory, No. 619, Jicui Street, Tianfu New District, Chengdu, 610213, Sichuan, China.
Abstract
Deep learning has been extensively employed in the prediction of metamaterial properties. However, the multi-layer perceptron-kernelled methods lack interpretability and are highly dependent on large datasets, making the end-to-end mapping opaque and computationally expensive and hindering the exploration and application of physical mechanisms. To address these issues, the Kolmogorov-Arnold Operat…
Create a free account to keep reading
Free members get 10 full research views every month across publications, clinical trials, FDA clearances, adverse events, and NIH grants. No credit card required.
Want unlimited research access? See Pro plans