An orbital-based representation for accurate Quantum Machine Learning

TL;DR

This paper presents a new electronic structure-based representation for quantum machine learning that improves the prediction of various electronic properties across diverse chemical compounds using inexpensive ab initio calculations.

Contribution

It introduces an orbital-based representation that explicitly accounts for electronic structure changes, enhancing the accuracy and flexibility of QML models for multiple properties.

Findings

Accurately predicts total potential energy, HOMO/LUMO energies, ionization potential, and electron affinity.

Demonstrates applicability to molecules with different charges and spin states.

Uses computationally inexpensive ab initio calculations for representation.

Abstract

We introduce an electronic structure based representation for quantum machine learning (QML) of electronic properties throughout chemical compound space. The representation is constructed using computationally inexpensive ab initio calculations and explicitly accounts for changes in the electronic structure. We demonstrate the accuracy and flexibility of resulting QML models when applied to property labels such as total potential energy, HOMO and LUMO energies, ionization potential, and electron affinity, using as data sets for training and testing entries from the QM7b, QM7b-T, QM9, and LIBE libraries. For the latter, we also demonstrate the ability of this approach to account for molecular species of different charge and spin multiplicity, resulting in QML models that infer total potential energies based on geometry, charge, and spin as input.