Construction of reactive potential energy surfaces with Gaussian process regression: active data selection

ABSTRACT

Gaussian process regression (GPR) is an efficient non-parametric method for constructing multi-dimensional potential energy surfaces (PESs) for polyatomic molecules. Since not only the posterior mean but also the posterior variance can be easily calculated, GPR provides a well-established model for active learning, through which PESs can be constructed more efficiently and accurately. We propose a strategy of active data selection for the construction of PESs with emphasis on low energy regions. Through three-dimensional (3D) example of H₃, the validity of this strategy is verified. The PESs for two prototypically reactive systems, namely, H + H₂O ↔ H₂ + OH reaction and H + CH₄ ↔ H₂ + CH₃ reaction are reconstructed. Only 920 and 4000 points are assembled to reconstruct these two PESs respectively. The accuracy of the GP PESs is not only tested by energy errors but also validated by quantum scattering calculations.

KEYWORDS:

• Potential energy surface
• Gaussian process regression
• active learning

Acknowledgments

The authors thank Rodrigo Vargas and Roman Krems for the update on machine learning. This work was supported by the National Natural Science Foundation of China (Grant Nos. 21688102, 21590800, 21433009), and the Strategic Priority Research Program of the Chinese Academy of Sciences (Grant No. XDB17010000).

Disclosure statement

No potential conflict of interest was reported by the authors.

Additional information

Funding

National Natural Science Foundation of China [grant number 21688102], [grant number 21590800], [grant number 21433009]; Strategic Priority Research Program of the Chinese Academy of Sciences [grant number XDB17010000].