New method to analyze short-range order in metallic glass models

Jason has published a paper using machine learning to discover short-range order in computer models of metallic glasses. The methods extracts SRO clusters from the model, aligns them using a ML method called point pattern matching (described in greater detail here), then uses ML clustering to group them together into prototypical structures. The code for executing the analysis is available on GitHub, and all the data for this paper is available on the Materials Data Facility.