A procedure is proposed to produce intermolecular potential energy surfaces from limited data. The procedure involves generation of geometrical configurations using a Latin hypercube design, with a maximin criterion, based on inverse internuclear distances. Gaussian processes are used to interpolate the data, using over-specified inverse molecular distances as covariates, greatly improving the interpolation. Symmetric covariance functions are specified so that the interpolation surface obeys all relevant symmetries, reducing prediction errors. The interpolation scheme can be applied to many important molecular interactions with trivial modifications. Results are presented for three systems involving , a system with a deep energy minimum (), and a system with 48 symmetries (). In each case, the procedure accurately predicts an independent test set. Training this method with high-precision evaluations of the interaction enables a parameter-free, first-principles prediction of the cross virial coefficient that agrees very well with experiments.
REFERENCES
This can be obtained from the maximum energy, before applying the energy cutoff over the test data (if available) or training data.
