Page 198 - Read Online

Review | Open Access

Journal of Materials Informatics

Wang et al. J. Mater. Inf. 2026, 6, 16 DOI:10.20517/jmi.2025.85

Multi-fidelity learning in materials informatics:
methodologies, applications, and outlook

Bo Wang , Yangyang Xu , Yumei Zhou , Dezhen Xue 1
1
2
1
Keywords:
Multi-fidelity learning,
materials informatics,
surrogate modeling, active
learning, data-driven
materials design
Citation: Wang, B.; Xu, Y.;
Zhou, Y.; Xue, D.
Multi-fidelity learning in
materials informatics:
methodologies, applications,
and outlook. J. Mater. Inf.
2026, 6, 16.
https://dx.doi.org/10.20517
/jmi.2025.85
Received: 8 Oct 2025
First Decision: 10 Nov Abstract
2025
Revised: 24 Dec 2025 Data-driven methods are transforming materials design by accelerating the discovery of
Accepted: 14 Jan 2026 new compounds and the optimization of existing systems. However, the progress of such
Published: 30 Mar 2026 approaches is often constrained by the scarcity of high-fidelity data from experiments and
advanced simulations. Multi-fidelity (MF) learning has emerged as a powerful strategy to
Academic Editor:
Zhimei Sun address this challenge by integrating information from diverse data sources that vary in
Copy Editor: accuracy and cost. In this review, we provide a systematic overview of the major
Ting-Ting Hu methodologies for MF learning, including statistical and parametric models, machine
Production Editor:
Ting-Ting Hu learning models with fidelity features, correction-based models such as co-kriging, deep
learning frameworks, and active learning frameworks. We discuss the strengths,
limitations, and typical applications of each method in materials science, with illustrative
examples spanning electronic structure modeling, alloy design, and interatomic potential
development. Cross-cutting issues are also examined, including the bias-variance trade-off,
data requirements for nested vs. non-nested designs, and computational scalability. Finally,
we highlight outstanding challenges and outline emerging opportunities, such as
physics-informed and generative MF models, standardized datasets, and integration with
autonomous laboratories. Together, these perspectives define a roadmap for advancing MF
learning as a core enabler of next-generation materials discovery.

1 State Key Laboratory for Mechanical Behavior of Materials, Xi’an Jiaotong University, Xi’an 710049, Shaanxi, China.
2 MOE Key Laboratory for Nonequilibrium Synthesis and Modulation of Condensed Matter, School of Physics, Xi’an Jiaotong University,
Xi’an 710049, Shaanxi, China.
Correspondence to: Dr. Yangyang Xu, MOE Key Laboratory for Nonequilibrium Synthesis and Modulation of Condensed Matter, School of
Physics, Xi’an Jiaotong University, Xi’an 710049, Shaanxi, China. E-mail: xuyangyang@xjtu.edu.cn; Prof. Yumei Zhou, Prof. Dezhen Xue,
State Key Laboratory for Mechanical Behavior of Materials, Xi’an Jiaotong University, Xi’an 710049, Shaanxi, China. E-mail:
zhouyumei@xjtu.edu.cn; xuedezhen@xjtu.edu.cn

www.oaepublish.com Submit a Manuscript: https://ucenter.oaepublish.com