Skip to main content
OSTI.GOVU.S. Department of Energy Office of Scientific and Technical Information
U.S. Department of Energy
Office of Scientific and Technical Information
 

Active learning for SNAP interatomic potentials via Bayesian predictive uncertainty

Journal Article · · Computational Materials Science
 [1];  [1];  [2];  [1]
  1. Sandia National Laboratories (SNL-CA), Livermore, CA (United States)
  2. Sandia National Laboratories (SNL-NM), Albuquerque, NM (United States)
+ Show Author Affiliations

Bayesian inference with a simple Gaussian error model is used to efficiently compute prediction variances for energies, forces, and stresses in the linear SNAP interatomic potential. Here, the prediction variance is shown to have a strong correlation with the absolute error over approximately 24 orders of magnitude. Using this prediction variance, an active learning algorithm is constructed to iteratively train a potential by selecting the structures with the most uncertain properties from a pool of candidate structures. The relative importance of the energy, force, and stress errors in the objective function is shown to have a strong impact upon the trajectory of their respective net error metrics when running the active learning algorithm. Batched training of different batch sizes is also tested against singular structure updates, and it is found that batches can be used to significantly reduce the number of retraining steps required with only minor impact on the active learning trajectory.

Research Organization:
Sandia National Laboratories (SNL-CA), Livermore, CA (United States)
Sponsoring Organization:
USDOE Office of Science (SC), Advanced Scientific Computing Research (ASCR); USDOE Office of Science (SC), Fusion Energy Sciences (FES)
Grant/Contract Number:
NA0003525
OSTI ID:
2372952
Report Number(s):
SAND--2024-07432J
Journal Information:
Computational Materials Science, Journal Name: Computational Materials Science Vol. 242; ISSN 0927-0256
Publisher:
ElsevierCopyright Statement
Country of Publication:
United States
Language:
English

References (32)

Active Learning in Bayesian Neural Networks for Bandgap Predictions of Novel Van der Waals Heterostructures journal August 2021
Uncertainty Quantification in Atomistic Modeling of Metals and Its Effect on Mesoscale and Continuum Modeling: A Review journal October 2020
Active learning of linearly parametrized interatomic potentials journal December 2017
Batch active learning for accelerating the development of interatomic potentials journal June 2022
Spectral neighbor analysis method for automated generation of quantum-accurate interatomic potentials journal March 2015
Gaussian Process Regression for Materials and Molecules journal August 2021
Explicit Multielement Extension of the Spectral Neighbor Analysis Potential for Chemically Complex Systems journal May 2020
Performance and Cost Assessment of Machine Learning Interatomic Potentials journal October 2019
On-the-Fly Active Learning of Interatomic Potentials for Large-Scale Atomistic Simulations journal July 2020
Efficient Atomic-Resolution Uncertainty Estimation for Neural Network Potentials Using a Replica Ensemble journal June 2020
Permutationally Invariant Polynomial Basis for Molecular Energy Surface Fitting via Monomial Symmetrization journal November 2009
Training data selection for accuracy and transferability of interatomic potentials journal September 2022
Automated calculation of thermal rate coefficients using ring polymer molecular dynamics and machine-learning interatomic potentials with active learning journal January 2018
Support vector machine regression (LS-SVM)—an alternative to artificial neural networks (ANNs) for the analysis of quantum chemistry data? journal January 2011
Machine learning of molecular properties: Locality and active learning journal June 2018
Extending the accuracy of the SNAP interatomic potential form journal June 2018
Less is more: Sampling chemical space with active learning journal June 2018
Automatic selection of atomic fingerprints and reference configurations for machine-learning potentials journal June 2018
Ring polymer molecular dynamics and active learning of moment tensor potential for gas-phase barrierless reactions: Application to S + H 2 journal December 2019
Committee neural network potentials control generalization errors and enable active learning journal September 2020
Uncertainty estimation for molecular dynamics and sampling journal February 2021
Fast uncertainty estimates in deep learning interatomic potentials journal April 2023
Accelerating crystal structure prediction by machine-learning interatomic potentials with active learning journal February 2019
Data-driven material models for atomistic simulation journal May 2019
Quality of uncertainty estimates from neural network potential ensembles journal January 2022
Gaussian Approximation Potentials: The Accuracy of Quantum Mechanics, without the Electrons journal April 2010
Fast and Accurate Modeling of Molecular Atomization Energies with Machine Learning journal January 2012
Finding Density Functionals with Machine Learning journal June 2012
Bayesian Ensemble Approach to Error Estimation of Interatomic Potentials journal October 2004
“Learn on the Fly”: A Hybrid Classical and Quantum-Mechanical Molecular Dynamics Simulation journal October 2004
Moment Tensor Potentials: A Class of Systematically Improvable Interatomic Potentials journal January 2016
FitSNAP: Atomistic machine learning with LAMMPS journal April 2023

Similar Records

Extending the accuracy of the SNAP interatomic potential form
Journal Article · Wed Mar 28 00:00:00 EDT 2018 · Journal of Chemical Physics · OSTI ID:1429723
Active Learning Framework
Software · Tue Jan 03 19:00:00 EST 2023 · OSTI ID:code-115537

Related Subjects

97 MATHEMATICS AND COMPUTING
Active learning
Bayesian inference
Interatomic potentials
Uncertainty quantification