Deep Learning Models to Identify Common Phases Across Material Systems from X-ray Diffraction

Datasets of synthesized x-ray diffraction patterns corresponding to mixed-phase samples from binary metal alloy material systems spanning 23 elements. Each of the two datasets is based on crystal structures from a different source: (1) experimentally measured from the Inorganic Crystal Structure Database and (2) theoretically predicted from the Materials Project. These datasets were used in our work under review, "Deep Learning Models to Identify Common Phases Across Material Systems from X-ray Diffraction," to train and assess convolutional neural network models to rapidly identify samples containing large fractions of A15-type phases.