Complete and Efficient Covariants for 3D Point Configurations with Application to Learning Molecular Quantum Properties

TL;DR

This paper develops complete and efficient $SO(3)$-covariant features for 3D point configurations, reducing computational complexity and enhancing modeling of molecular quantum properties.

Contribution

It introduces a complete set of higher order covariant features for 3D points and simplifies their computation, improving molecular property modeling.

Findings

$6k-5$ features suffice for up to $k$ atoms.

Replaced Clebsch--Gordan operations with matrix multiplications.

Reduced feature computation complexity from $O(l^6)$ to $O(l^3)$.

Abstract

When modeling physical properties of molecules with machine learning, it is desirable to incorporate $SO(3)$-covariance. While such models based on low body order features are not complete, we formulate and prove general completeness properties for higher order methods, and show that $6k-5$ of these features are enough for up to $k$ atoms. We also find that the Clebsch--Gordan operations commonly used in these methods can be replaced by matrix multiplications without sacrificing completeness, lowering the scaling from $O(l^6)$ to $O(l^3)$ in the degree of the features. We apply this to quantum chemistry, but the proposed methods are generally applicable for problems involving 3D point configurations.