Predicting Properties of Periodic Systems from Cluster Data: A Case Study of Liquid Water
arXiv:2312.01414 · doi:10.1063/5.0078983
Abstract
The accuracy of the training data limits the accuracy of bulk properties from machine-learned potentials. For example, hybrid functionals or wave-function-based quantum chemical methods are readily available for cluster data but effectively out-of-scope for periodic structures. We show that local, atom-centred descriptors for machine-learned potentials enable the prediction of bulk properties from cluster model training data, agreeing reasonably well with predictions from bulk training data. We demonstrate such transferability by studying structural and dynamical properties of bulk liquid water with density functional theory and have found an excellent agreement with experimental as well as theoretical counterparts.
References in corpus (11)
- How van der Waals interactions determine the unique properties of water
- SchNet: A continuous-filter convolutional neural network for modeling quantum interactions
- The Individual and Collective Effects of Exact Exchange and Dispersion Interactions on the Ab Initio Structure of Liquid Water
- An Accurate and Transferable Machine Learning Potential for Carbon
- Gaussian Moments as Physically Inspired Molecular Descriptors for Accurate and Scalable Machine Learning Potentials
- Uncertainty estimation for molecular dynamics and sampling
- PiNN: A Python Library for Building Atomic Neural Networks of Molecules and Materials
- Comparing the Accuracy of High-Dimensional Neural Network Potentials and the Systematic Molecular Fragmentation Method: A Benchmark Study for all-trans Alkanes
- A Novel Approach to Describe Chemical Environments in High Dimensional Neural Network Potentials
- Fast and Sample-Efficient Interatomic Neural Network Potentials for Molecules and Materials Based on Gaussian Moments
- Neural-Network Assisted Study of Nitrogen Atom Dynamics on Amorphous Solid Water. I. Adsorption & Desorption
Cited by in corpus (3)
- Transfer learning for chemically accurate interatomic neural network potentials
- Reaction dynamics on amorphous solid water surfaces using interatomic machine learned potentials. Microscopic energy partition revealed from the P + H -> PH reaction
- CO Adsorption Sites on Interstellar Water Ices Explored with Machine Learning Potentials. Binding energy distributions and snowline