Abstract
Determination of crystal system and space group in the initial stages of crystal structure analysis forms a bottleneck in material science workflow that often requires manual tuning. Herein we propose a machine-learning (ML)-based approach for crystal system and space group classification based on powder X-ray diffraction (XRD) patterns as a proof of concept using simulated patterns. Our tree-ensemble-based ML model works with nearly or over 90% accuracy for crystal system classification, except for triclinic cases, and with 88% accuracy for space group classification with five candidates. We also succeeded in quantifying empirical knowledge vaguely shared among experts, showing the possibility for data-driven discovery of unrecognised characteristics embedded in experimental data by using an interpretable ML approach.
Highlights
Determination of crystal system and space group in the initial stages of crystal structure analysis forms a bottleneck in material science workflow that often requires manual tuning
We show that a simple and fast ML technique (Fig. 1) can classify crystal systems and space groups (230 classes) with high accuracy based on powder X-ray diffraction (XRD) patterns and that data-driven quantification of empirical expert knowledge is possible using an interpretable ML model
Data preparation and feature extraction. 199,391 powder XRD patterns were calculated as training datasets from Inorganic Crystal Structure Database (ICSD) entries using Pymatgen middleware
Summary
Determination of crystal system and space group in the initial stages of crystal structure analysis forms a bottleneck in material science workflow that often requires manual tuning. Park et al classified crystal systems and space groups by applying a convolutional neural network (CNN) to simulated powder XRD patterns. They achieved high classification performance despite data deterioration due to Poisson noise and instrumental resolution. We show that a simple and fast ML technique (Fig. 1) can classify crystal systems (seven classes) and space groups (230 classes) with high accuracy based on powder XRD patterns and that data-driven quantification of empirical expert knowledge is possible using an interpretable ML model. This study is in a proof of concept (POC) stage using ML models trained on ideal simulated diffraction data (i.e. noise-free, no peak-broadening, no peak superposition, no impurity peaks), it provided some interesting findings described in following sections
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a callDisclaimer: All third-party content on this website/platform is and will remain the property of their respective owners and is provided on "as is" basis without any warranties, express or implied. Use of third-party content does not indicate any affiliation, sponsorship with or endorsement by them. Any references to third-party content is to identify the corresponding services and shall be considered fair use under The CopyrightLaw.
