Robust and scalable uncertainty estimation with conformal prediction for machine-learned interatomic potentials

被引:11
|
作者
Hu, Yuge [1 ]
Musielewicz, Joseph [2 ]
Ulissi, Zachary W. [2 ]
Medford, Andrew J. [1 ]
机构
[1] Georgia Inst Technol, Dept Chem & Biomol Engn, Atlanta, GA 30332 USA
[2] Carnegie Mellon Univ, Dept Chem Engn, Pittsburgh, PA USA
来源
关键词
machine learning force fields; neural network force fields; uncertainty quantification; conformal prediction; DENSITY-FUNCTIONAL THEORY;
D O I
10.1088/2632-2153/aca7b1
中图分类号
TP18 [人工智能理论];
学科分类号
081104 ; 0812 ; 0835 ; 1405 ;
摘要
Uncertainty quantification (UQ) is important to machine learning (ML) force fields to assess the level of confidence during prediction, as ML models are not inherently physical and can therefore yield catastrophically incorrect predictions. Established a-posteriori UQ methods, including ensemble methods, the dropout method, the delta method, and various heuristic distance metrics, have limitations such as being computationally challenging for large models due to model re-training. In addition, the uncertainty estimates are often not rigorously calibrated. In this work, we propose combining the distribution-free UQ method, known as conformal prediction (CP), with the distances in the neural network's latent space to estimate the uncertainty of energies predicted by neural network force fields. We evaluate this method (CP+latent) along with other UQ methods on two essential aspects, calibration, and sharpness, and find this method to be both calibrated and sharp under the assumption of independent and identically-distributed (i.i.d.) data. We show that the method is relatively insensitive to hyperparameters selected, and test the limitations of the method when the i.i.d. assumption is violated. Finally, we demonstrate that this method can be readily applied to trained neural network force fields with traditional and graph neural network architectures to obtain estimates of uncertainty with low computational costs on a training dataset of 1 million images to showcase its scalability and portability. Incorporating the CP method with latent distances offers a calibrated, sharp and efficient strategy to estimate the uncertainty of neural network force fields. In addition, the CP approach can also function as a promising strategy for calibrating uncertainty estimated by other approaches.
引用
收藏
页数:23
相关论文
共 50 条
  • [1] How to validate machine-learned interatomic potentials
    Morrow, Joe D.
    Gardner, John L. A.
    Deringer, Volker L.
    [J]. JOURNAL OF CHEMICAL PHYSICS, 2023, 158 (12):
  • [2] A FRAMEWORK FOR A GENERALIZATION ANALYSIS OF MACHINE-LEARNED INTERATOMIC POTENTIALS
    Ortner, Christoph
    Wang, Yangshuai
    [J]. MULTISCALE MODELING & SIMULATION, 2023, 21 (03): : 1053 - 1080
  • [3] Simple machine-learned interatomic potentials for complex alloys
    Byggmastar, J.
    Nordlund, K.
    Djurabekova, F.
    [J]. PHYSICAL REVIEW MATERIALS, 2022, 6 (08)
  • [4] Machine-learned interatomic potentials: Recent developments and prospective applications
    Eyert, Volker
    Wormald, Jonathan
    Curtin, William A.
    Wimmer, Erich
    [J]. JOURNAL OF MATERIALS RESEARCH, 2023, 38 (24) : 5079 - 5094
  • [5] Machine-learned interatomic potentials for alloys and alloy phase diagrams
    Rosenbrock, Conrad W.
    Gubaev, Konstantin
    Shapeev, Alexander V.
    Partay, Livia B.
    Bernstein, Noam
    Csanyi, Gabor
    Hart, Gus L. W.
    [J]. NPJ COMPUTATIONAL MATERIALS, 2021, 7 (01)
  • [6] Machine-learned interatomic potentials: Recent developments and prospective applications
    Volker Eyert
    Jonathan Wormald
    William A. Curtin
    Erich Wimmer
    [J]. Journal of Materials Research, 2023, 38 : 5079 - 5094
  • [7] Machine-learned interatomic potentials for alloys and alloy phase diagrams
    Conrad W. Rosenbrock
    Konstantin Gubaev
    Alexander V. Shapeev
    Livia B. Pártay
    Noam Bernstein
    Gábor Csányi
    Gus L. W. Hart
    [J]. npj Computational Materials, 7
  • [8] Machine-learned interatomic potentials by active learning: amorphous and liquid hafnium dioxide
    Sivaraman, Ganesh
    Krishnamoorthy, Anand Narayanan
    Baur, Matthias
    Holm, Christian
    Stan, Marius
    Csanyi, Gabor
    Benmore, Chris
    Vazquez-Mayagoitia, Alvaro
    [J]. NPJ COMPUTATIONAL MATERIALS, 2020, 6 (01)
  • [9] Machine-learned interatomic potentials by active learning: amorphous and liquid hafnium dioxide
    Ganesh Sivaraman
    Anand Narayanan Krishnamoorthy
    Matthias Baur
    Christian Holm
    Marius Stan
    Gábor Csányi
    Chris Benmore
    Álvaro Vázquez-Mayagoitia
    [J]. npj Computational Materials, 6
  • [10] Constructing and Evaluating Machine-Learned Interatomic Potentials for Li-Based Disordered Rocksalts
    Choyal, Vijay
    Sagar, Nidhish
    Sai Gautam, Gopalakrishnan
    [J]. JOURNAL OF CHEMICAL THEORY AND COMPUTATION, 2024, 20 (11) : 4844 - 4856