Data-driven learning of 3-point correlation functions as microstructure representations

This paper considers the open challenge of identifying complete, concise, and explainable quantitative microstructure representations for disordered heterogeneous material systems. Completeness conciseness have been achieved through existing data-driven methods, e.g., deep generative models, which, however, do not provide mathematically latent representations. study investigates composed three-point correlation functions, which are a special type spatial convolutions. We show that variety microstructures can be characterized by concise subset correlations (100-fold smaller than full set), identification such subsets Bayesian optimization on small dataset. The proposed representation directly used to compute properties leveraging effective medium theory, allowing construction predictive structure-property models with significantly less data needed purely methods computational cost 100-fold lower physics-based model.