Machine learning-based structure–property predictions in silica aerogels

An artificial neural network is developed to predict the fractal properties of silica aerogels, modelled via diffusion-limited cluster–cluster aggregation, and then inverted for reconstructing an optimised network for a target fractal dimension.