Imagen de Portada

Learning Self-Supervised Representations of Powder-Diffraction Patterns

Guardado en:
Publicado en:Crystals vol. 15, no. 5 (2025), p. 393
Autor principal: Das Shubhayu
Otros Autores: Vorholt, Markus, Houben, Andreas, Dronskowski, Richard
Publicado:
MDPI AG
Materias:
Crystal structure
X ray powder diffraction
Machine learning
Accuracy
Self-supervised learning
Neural networks
Classification
X-ray diffraction
Crystallography
Algorithms
Diffraction patterns
Representations
Extinction
Acceso en línea:Citation/Abstract
Full Text + Graphics
Full Text - PDF
Etiquetas: Agregar Etiqueta
Sin Etiquetas, Sea el primero en etiquetar este registro!
Resumen:The potential of machine learning (ML) models for predicting crystallographic symmetry information from single-phase powder X-ray diffraction (XRD) patterns is investigated. Given the scarcity of large, labeled experimental datasets, we train our models using simulated XRD patterns generated from crystallographic databases. A key challenge in developing reliable diffraction-based structure-solution tools lies in the limited availability of training data and the presence of natural adversarial examples, which hinder model generalization. To address these issues, we explore multiple training pipelines and testing strategies, including evaluations on experimental XRD data. We introduce a contrastive representation learning approach that significantly outperforms previous supervised learning models in terms of robustness and generalizability, demonstrating improved invariance to experimental effects. These results highlight the potential of self-supervised learning in advancing ML-driven crystallographic analysis.
ISSN:2073-4352
DOI:10.3390/cryst15050393
Fuente:Materials Science Database