Machine Learning Based Classification of crystal system using rendered images from X-ray diffraction (XRD) dataset

Abstract

X-ray diffraction(XRD) is an essential characterization technique to study the properties of the materials. Finding a material's crystal system is an important step in its analysis. So, the process should be fast as well as accurate. In total there are seven crystal systems: triclinic, monoclinic, orthorhombic, tetragonal, trigonal, hexagonal, and cubic. Previous studies have worked on finding the material crystal structure by introducing machine learning approaches. In, recent studies the X-ray diffraction(XRD) dataset in the tabular form was used to train a machine learning model to classify the material's crystal system. The machine learning models trained on the tabular X-ray diffraction(XRD) dataset didn't maximize their performance. In the scope of this study rendered X-ray diffraction(XRD) images had been used to maximize the performance of the machine learning models. By using the rendered images as input from the X-ray diffraction(XRD) tabular dataset the machine learning model was able to achieve an accuracy of 98%-99%. The final findings had shown that rendered images datasets had improved the machine learning model's ability to correctly classify the crystal systems of materials.