Skip to content

Latest commit

 

History

History
12 lines (8 loc) · 977 Bytes

README.md

File metadata and controls

12 lines (8 loc) · 977 Bytes

Bayesian Force Field Optimization with Local Gaussian Process Molecular Dynamics (LGPMD)

General Information

Source code for constructing local Gaussian process surrogate models for accelerated Bayesian force field optimization. The code is adaptable to any quantity-of-interest in chemistry in physics applications with appropriate updates to training and test data matrices as well as Gaussian process mean and kernel functions.

Required Software

Version History

LGPMD v1.0 - Local Gaussian process surrogate model for radial distribution functions of monatomic fluids. Hyperparameter training performed using leave-one-out log marginal likelihood maximization over a fixed hyperparameter space.

Acknowledgement

The source code development was supported by the National Science Foundation under award number CBET-1847340. Developed at the University of Utah, Department of Chemical Engineering by Brennon Shanks, Harry Sullivan and Michael P. Hoepfner.