Machine learning real space microstructure characteristics from scattering data.

Abstract

Using tools from morphological image analysis, we characterise spinodal decomposition microstructures by their Minkowski functionals, and search for a correlation between them and data from scattering experiments. To do this, we employ machine learning in the form of Gaussian process regression on data derived from numerical simulations of spinodal decomposition in polymer blends. For a range of microstructures, we analyse the predictions of the Minkowski functionals achieved by four Gaussian process regression models using the scattering data. Our findings suggest that there is a strong correlation between the scattering data and the Minkowski functionals.