Pre-Refinement Prediction of Dopant Site Occupancy in M-Type Hexaferrites via Hierarchical and Uncertainty-Aware Machine-Learning.
5 affiliations
- 1Department of Physics, Beijing Technology and Business University, Beijing, China.
- 2Ningbo Institute of Materials Technology and Engineering, Chinese Academy of Sciences, Ningbo, China.
- 3Feicui Lake Laboratory, School of Materials Science and Engineering, Anhui University, Hefei, China.
- 4Spallation Neutron Source Science Center (SNSSC), Dongguan, China.
- 5State Key Laboratory of Magnetism, Institute of Physics, Chinese Academy of Sciences, Beijing, China.
Abstract
Dopant site occupancy bridges chemical composition, local crystal structure, and macroscopic functional responses in M-type hexaferrites. However, reliable occupancy hypotheses are difficult to establish before structural refinement. Here, we develop a pre-refinement machine-learning framework using 494 dopant-level records extracted from refined structures. This workflow combines Ba/Sr-vs.-Fe sublattice classification, fractional occupancy prediction over five inequivalent Fe sites, and reliability-aware assessment. The CatBoost classifier achieves 96% accuracy on external validation for sublattice assignment. For Fe-sublattice dopants, homogeneous occupancy-model screening and heterogeneous cooperative occupancy-model screening yield a macro mean absolute error as low as 0.0584, while preserving dominant-site identification and occupancy ranking. The results show that dopant site occupancy is jointly controlled by dopant chemistry, target-dopant amount, and co-doping environment rather than a universal site-preference rule. By integrating TabPFN baseline comparison, training-experience assessment, and model-consistency analysis, the framework identifies boundary and extrapolative samples that require priority refinement. Our predictions for W-doped and Sc-doped M-type hexaferrites successfully reproduce key experimental occupancy trends. This work offers a practical data-driven strategy to generate physically plausible occupancy hypotheses and assign refinement priorities for partially occupied oxide systems.
Related
- Uncertainty-Driven Self-Refinement of Adaptive Kolmogorov-Arnold Networks for Fitting Nonadiabatic Potential Energy Surface
- Dataset of magnetic properties, structure and synthesis of doped barium and strontium M-type hexaferrites
- Anisotropy-controlled X-band microwave attenuation in Co–Ti Co-substituted Ba0.5Sr0.5Y M-type hexaferrites
- Publication-block calibration reveals wide uncertainty in PLGA release endpoint prediction
- NephroTox AI: an explainable machine learning framework with uncertainty quantification for drug-induced nephrotoxicity prediction
BibTeX
@article{liu2026prerefinement,
title = {{Pre-Refinement Prediction of Dopant Site Occupancy in M-Type Hexaferrites via Hierarchical and Uncertainty-Aware Machine-Learning.}},
author = {Dan Liu and Jianyu Ren and Ruoshui Liu and Yinong Yin and Jianfeng Xi and Xin Ming and Jiazheng Hao and Tongyun Zhao and Fengxia Hu and Baogen Shen},
journal = {Advanced science (Weinheim, Baden-Wurttemberg, Germany)},
year = {2026},
month = oct,
doi = {10.1002/advs.78493},
url = {https://europepmc.org/article/MED/42958698},
note = {CC BY},
}
RIS
TY - JOUR
TI - Pre-Refinement Prediction of Dopant Site Occupancy in M-Type Hexaferrites via Hierarchical and Uncertainty-Aware Machine-Learning.
AU - Dan Liu
AU - Jianyu Ren
AU - Ruoshui Liu
AU - Yinong Yin
AU - Jianfeng Xi
AU - Xin Ming
AU - Jiazheng Hao
AU - Tongyun Zhao
AU - Fengxia Hu
AU - Baogen Shen
PY - 2026
DA - 2026/10/29
JO - Advanced science (Weinheim, Baden-Wurttemberg, Germany)
DO - 10.1002/advs.78493
UR - https://europepmc.org/article/MED/42958698
AB - Dopant site occupancy bridges chemical composition, local crystal structure, and macroscopic functional responses in M-type hexaferrites. However, reliable occupancy hypotheses are difficult to establish before structural refinement. Here, we develop a pre-refinement machine-learning framework using 494 dopant-level records extracted from refined structures. This workflow combines Ba/Sr-vs.-Fe sublattice classification, fractional occupancy prediction over five inequivalent Fe sites, and reliability-aware assessment. The CatBoost classifier achieves 96% accuracy on external validation for sublattice assignment. For Fe-sublattice dopants, homogeneous occupancy-model screening and heterogeneous cooperative occupancy-model screening yield a macro mean absolute error as low as 0.0584, while preserving dominant-site identification and occupancy ranking. The results show that dopant site occupancy is jointly controlled by dopant chemistry, target-dopant amount, and co-doping environment rather than a universal site-preference rule. By integrating TabPFN baseline comparison, training-experience assessment, and model-consistency analysis, the framework identifies boundary and extrapolative samples that require priority refinement. Our predictions for W-doped and Sc-doped M-type hexaferrites successfully reproduce key experimental occupancy trends. This work offers a practical data-driven strategy to generate physically plausible occupancy hypotheses and assign refinement priorities for partially occupied oxide systems.
SN - 2198-3844
ER -