Mechanical Property Prediction in High‐Entropy Ceramics with Sparse Data via Diffusion‐CVAE‐Enhanced Neural Networks
Journal of the American Ceramic Society · 2026-07 · 109 (7)
DOI: 10.1111/jace.70994 ↗Kurz gefasst
Fachartikel in der Zeitschrift Journal of the American Ceramic Society, online erschienen am 21.07.2026. Verfasst von Shurui Shi und 4 weiteren Autor:innen. Die Arbeit nennt 27 Literaturangaben.
Abstract (Originalsprache)
High‐entropy carbide ceramics (HECCs) show great potential for extreme environments due to their excellent mechanical properties and thermal stability; however, the vast compositional space makes experimental exploration and first‐principles calculations costly. In this study, three machine learning models, random forest (RF), support vector regression (SVR), and artificial neural network (ANN), were developed to predict the elastic constants C 11 , C 12 , and C 44 of cubic HECCs. To address limited data availability, a data augmentation strategy based on Diffusion‐CVAE was introduced, which significantly improved model performance in most tasks. A database containing 126 119 HECC compositions was constructed, and the trained ANN model was used for high‐throughput prediction of elastic constants. Selected systems were further validated using density functional theory (DFT), showing strong agreement with the predictions. These results demonstrate that the proposed model achieves high accuracy and robustness, effectively capturing the relationship between composition and elastic properties, and provides an efficient data‐driven approach for rapid property prediction and compositional design of HECCs.
Zitieren
APA
Shi, S., Wang, D., Ding, Z., Zhao, B., & Wei, S. (2026). Mechanical Property Prediction in High‐Entropy Ceramics with Sparse Data via Diffusion‐CVAE‐Enhanced Neural Networks. Journal of the American Ceramic Society, 109(7). https://doi.org/10.1111/jace.70994
Vancouver
Shi S, Wang D, Ding Z, Zhao B, Wei S. Mechanical Property Prediction in High‐Entropy Ceramics with Sparse Data via Diffusion‐CVAE‐Enhanced Neural Networks. Journal of the American Ceramic Society. 2026;109(7). doi:10.1111/jace.70994
BibTeX
@article{Shi2026,
author = {Shi, Shurui and Wang, Dong and Ding, Zhengwei and Zhao, Bo and Wei, Shuhang},
title = {Mechanical Property Prediction in High‐Entropy Ceramics with Sparse Data via Diffusion‐CVAE‐Enhanced Neural Networks},
journal = {Journal of the American Ceramic Society},
year = {2026},
volume = {109},
number = {7},
doi = {10.1111/jace.70994}
}
EUVIMED