2020/11/30 by Apurba Nandi, Chen Qu, Paul Houston +2 · 1 citation
Physics and Astronomy · #physics.chem-ph
paper · pdf · doi:10.1063/5.0038301
published as J. Chem. Phys. 154 (2021) 051102
arxiv created 2021/05/17 · arxiv updated 2026/08/03
``Δ-machine learning" refers to a machine learning approach to bring a property such as a potential energy surface (PES) based on low-level (LL) density functional theory (DFT) energies and gradients to close to a coupled cluster (CC) level of accuracy. Here we present such an approach that uses the permutationally invariant polynomial (PIP) method to fit high-dimensional PESs. The approach is represented by a simple equation, in obvious notation V_LL→CC=VLL+ΔVCC-LL, and demonstrated for \ceCH4, \ceH3O+, and trans and cis-N-methyl acetamide (NMA), \ceCH3CONHCH3. For these molecules, the LL PES, VLL, is a PIP fit to DFT/B3LYP/6-31+G(d) energies and gradients, and ΔVCC-LL is a precise PIP fit obtained using a low-order PIP basis set and based on a relatively small number of CCSD(T) energies. For \ceCH4 these are new calculations adopting an aug-cc-pVDZ basis, for \ceH3O+ previous CCSD(T)-F12/aug-cc-pVQZ energies are used, while for NMA new CCSD(T)-F12/aug-cc-pVDZ calculations are performed. With as few as 200 CCSD(T) energies, the new PESs are in excellent agreement with benchmark CCSD(T) results for the small molecules, and for 12-atom NMA training is done with 4696 CCSD(T) energies.