Improving accuracy of interatomic potentials: more physics or more data? A case study of silica

  title={Improving accuracy of interatomic potentials: more physics or more data? A case study of silica},
  author={Ivan S. Novikov and Alexander V. Shapeev},
  journal={Materials Today Communications},

Figures and Tables from this paper

A machine-learned interatomic potential for silica and its relation to empirical models
Silica (SiO 2 ) is an abundant material with a wide range of applications. Despite much progress, the atomistic modelling of the different forms of silica has remained a challenge. Here we show that
Machine-Learning Interatomic Potentials for Materials Science
Large-scale atomistic computer simulations of materials rely on interatomic potentials providing computationally efficient predictions of energy and Newtonian forces. Traditional potentials have
Uncertainty quantification in molecular simulations with dropout neural network potentials
A class of Dropout Uncertainty Neural Network (DUNN) potentials are proposed that provide rigorous uncertainty estimates that can be understood from both Bayesian and frequentist statistics perspectives and can serve as a predictor for the accuracy of a calculation.
Deep-learning interatomic potential for irradiation damage simulations in MoS2 with ab initial accuracy
Potentials that could accurately describe the irradiation damage processes are highly desired to figure out the atomic-level response of various newly-discovered materials under irradiation
The MLIP package: moment tensor potentials with MPI and active learning
This paper illustrates how to construct moment tensor potentials using active learning as implemented in the MLIP package, focusing on the efficient ways to sample configurations for the training set, how expanding theTraining set changes the error of predictions, how to set up ab initio calculations in a cost-effective manner, etc.
Machine-learned interatomic potentials for alloys and alloy phase diagrams
We introduce machine-learned potentials for Ag-Pd to describe the energy of alloy configurations over a wide range of compositions. We compare two different approaches. Moment tensor potentials
Making thermal rate constant calculations reliable using best practices: case study of OH + HBr $\to$ Br + H$_2$O
In the present work we apply the combination of Moment Tensor Potential (MTP) and Ring Polymer Molecular Dynamics (RPMD) to the calculation of the thermal rate constants of the OH + HBr → Br + H 2 O


Accelerating high-throughput searches for new alloys with active learning of interatomic potentials
Machine Learning Predictions of Molecular Properties: Accurate Many-Body Potentials and Nonlocality in Chemical Space
A systematic hierarchy of efficient empirical methods to estimate atomization and total energies of molecules and is achieved by a vectorized representation of molecules (so-called Bag of Bonds model) that exhibits strong nonlocality in chemical space.
Data-Driven Learning of Total and Local Energies in Elemental Boron.
This model readily provides atom-resolved, local energies and thus deepened insight into the frustrated β-rhombohedral boron structure, and opens the door for the efficient and automated generation of GAPs, and other machine-learning-based interatomic potentials, to suggest their usefulness as a tool for materials discovery.
Transferable Dynamic Molecular Charge Assignment Using Deep Neural Networks.
HIP-NN charge predictions are many orders of magnitude faster than direct DFT calculations, and combined results provide further evidence that ML (specifically HIP-NN) provides a pathway to greatly increase the range of feasible simulations while retaining quantum-level accuracy.
Representing the potential-energy surface of protonated water clusters by high-dimensional neural network potentials.
A reactive full-dimensional NN potential for protonated water clusters up to the octamer is presented, showing that the energetic, structural, and vibrational properties are in excellent agreement with DFT results making the NN approach a very promising candidate for developing a high-quality potential for water.
Machine learning based interatomic potential for amorphous carbon
We introduce a Gaussian approximation potential (GAP) for atomistic simulations of liquid and amorphous elemental carbon. Based on a machine learning representation of the density-functional theory
A first principles based polarizable O(N) interatomic force field for bulk silica.
A reformulation of the Tangney-Scandolo interatomic force field for silica is presented, which removes the requirement to perform an Ewald summation and the resulting O(N) scheme makes it possible to model hundreds of thousands of atoms with modest computational resources, without compromising the force field accuracy.
Learning scheme to predict atomic forces and accelerate materials simulations
It is argued that pathways are available to systematically and continuously improve the predictive capability of such a learned force field in an adaptive manner, and that this concept can be generalized to include multiple elements.
Symmetry-Adapted Machine Learning for Tensorial Properties of Atomistic Systems.
This work introduces a formalism that extends existing schemes and makes it possible to perform machine learning of tensorial properties of arbitrary rank, and for general molecular geometries, and derives a tensor kernel adapted to rotational symmetry.
Fast machine learning models of electronic and energetic properties consistently reach approximation errors better than DFT accuracy
Numerical evidence is presented that ML model predictions for all properties can reach an approximation error to DFT which is on par with chemical accuracy.