Generalizing an Energy Predictor based on Wavelet Scattering for 3D Atomic Systems

The dream of machine learning in quantum matter is for a neural network to learn the underlying physics of an atomic system, allowing it to move beyond interpolation of the training set to the prediction of properties that were not present in the original training data. Achieving this ambitious goal will require a method to convert a 3D atomic system into neural-network-friendly features that preserve rotational and translational symmetry, smoothness under small perturbations, and invariance under re-ordering. The atomic orbital wavelet scattering transform preserves these symmetries by construction, and has achieved great success as a featurization method for machine learning energy prediction. Both in small molecules and in the amorphous Li_xSi system, neural networks using wavelet scattering coefficients as features have demonstrated a comparable accuracy to Density Functional Theory at a small fraction of the computational cost. In this work, we test the generalizability of our Li_xSi energy predictor to properties that were not included in the training set, such as elastic constants and migration barriers. We also discuss the potential for future improvements in generalizability through automatic training-set expansion based on active learning.