Machine learning in neutron scattering data analysis

Neutron scattering is one of the state-of-the-art techniques for detecting the structural and dynamic properties of materials. The data analysis of neutron scattering is an inverse process that extracts hidden features from data and correlates them with information about the structure and properties of samples. With the global popularity of machine learning, its powerful automatic feature extraction capability was noticed by data analysis scientists. In recent years, the integration of neutron scattering data analysis and machine learning methods has seen significant development. In this paper, the applications of machine learning in data analysis for common neutron scattering techniques, including neutron diffraction, small angle neutron scattering, neutron reflectometry, and neutron imaging, were systematically reviewed. We classified this research into different neutron scattering techniques and different themes for each technique. Building upon the review, we discussed the application paradigms and current challenges associated with machine learning methods in neutron scattering data analysis.