Materials, Journal Year: 2025, Volume and Issue: 18(5), P. 959 - 959

Published: Feb. 21, 2025

This study proposes a novel framework using graph convolutional networks to analyze and interpret X-ray diffraction patterns, addressing challenges in phase identification for multi-phase materials. By representing patterns as graphs, the captures both local global relationships between peaks, enabling accurate even presence of overlapping peaks noisy data. The outperforms traditional machine learning models, achieving precision 0.990 recall 0.872. performance is attained with minimal hyperparameter tuning, making it scalable large-scale material discovery applications. Data augmentation, including synthetic data generation noise injection, enhances model’s robustness by simulating real-world experimental variations. However, reliance on computational cost construction inference remain limitations. Future work will focus integrating real data, optimizing efficiency, exploring lightweight architectures improve scalability high-throughput

Language: Английский