# Methods in PES-Learn: Direct-Fit Machine Learning of Born–Oppenheimer Potential Energy Surfaces

## TL;DR

PES-Learn 1.0 is an open-source tool for building machine learning models of molecular potential energy surfaces, with new features like a Python API and kernel ridge regression.

## Contribution

The introduction of kernel ridge regression and improved interoperability via a new Python API in PES-Learn.

## Key findings

- Kernel ridge regression is effective for modeling semi-global potential energy surfaces.
- PES-Learn's performance is benchmarked using benzene and ethanol datasets from the rMD17 database.
- Neural network models in PES-Learn can predict gradients for ethanol and benzene.

## Abstract

The release of PES-Learn version 1.0 as an open-source software package for the automatic construction of machine learning models of semi-global molecular potential energy surfaces (PESs) is presented. Improvements to PES-Learn’s interoperability are stressed with new Python API that simplifies workflows for PES construction via interaction with QCSchema input and output infrastructure. In addition, a new machine learning method is introduced to PES-Learn: kernel ridge regression (KRR). The capabilities of KRR are emphasized with examination of select semi-global PESs. All machine learning methods available in PES-Learn are benchmarked with benzene and ethanol datasets from the rMD17 database to illustrate PES-Learn’s performance ability. Fitting performance and timings are assessed for both systems. Finally, the ability to predict gradients with neural network models is presented and benchmarked with ethanol and benzene. PES-Learn is an active project and welcomes community suggestions and contributions.