Big Data meets Quantum Chemistry Approximations: The $\Delta$-Machine Learning Approach

TL;DR

This paper introduces a machine learning correction method that significantly enhances the accuracy of inexpensive quantum chemistry calculations, enabling large-scale, precise thermochemical and electronic property predictions.

Contribution

The authors develop a novel $ ext{ extDelta}$-machine learning approach that improves quantum chemistry predictions with transferability across different methods and large molecular datasets.

Findings

Achieves chemical accuracy for thermochemical properties of 16,000 molecules.

Predicts electron correlation energies at Hartree-Fock cost.

Demonstrates transferability to large molecular datasets.

Abstract

Chemically accurate and comprehensive studies of the virtual space of all possible molecules are severely limited by the computational cost of quantum chemistry. We introduce a composite strategy that adds machine learning corrections to computationally inexpensive approximate legacy quantum methods. After training, highly accurate predictions of enthalpies, free energies, entropies, and electron correlation energies are possible, for significantly larger molecular sets than used for training. For thermochemical properties of up to 16k constitutional isomers of C$_7$H$_{10}$O$_2$ we present numerical evidence that chemical accuracy can be reached. We also predict electron correlation energy in post Hartree-Fock methods, at the computational cost of Hartree-Fock, and we establish a qualitative relationship between molecular entropy and electron correlation. The transferability of our approach is demonstrated, using semi-empirical quantum chemistry and machine learning models trained on 1 and 10\% of 134k organic molecules, to reproduce enthalpies of all remaining molecules at density functional theory level of accuracy.