Automatically growing global reactive neural network potential energy surfaces: A trajectory-free active learning strategy

Lin, Qidong; Zhang, Yaolong; Zhao, Bin; Jiang, Bin

doi:10.1063/5.0004944

Cited by 54 publications

(69 citation statements)

References 56 publications

(81 reference statements)

Supporting

Mentioning

Contrasting

Order By: Relevance

“…In this sense, ML models have a huge potential to enhance MQCD simulations by providing the electronic PESs and enabling the investigation of reactions that are not feasible with conventional approaches. 15 , 175 , 434 , 435 In fact, most studies to date that describe photochemistry with ML aim to replace the quantum chemical calculation of the PESs in MQCD approaches.…”

Section: Quantum Chemical Theory and Methodsmentioning

confidence: 99%

“… 515 Noticeable, within ML for quantum chemistry, active learning often refers to an approach, where an initial training set is used to fit an ML model, and this previously learned information is applied to expand the training set. 435 The latter approach is often carried out with the help of MD simulations. Simulation of many trajectories on-the-fly and estimation of the reliability of the ML-fitted PESs at each time step are powerful to identify under-sampled or unknown regions of the PESs.…”

Section: Data Sets For Excited Statesmentioning

confidence: 99%

“…MD simulations would take a long time to overcome the potential barriers and reach the region of unknown molecular structures. 435 …”

Section: Data Sets For Excited Statesmentioning

confidence: 99%

See 2 more Smart Citations

Machine Learning for Electronically Excited States of Molecules

2020

View full text Add to dashboard Cite

Electronically excited states of molecules are at the heart of photochemistry, photophysics, as well as photobiology and also play a role in material science. Their theoretical description requires highly accurate quantum chemical calculations, which are computationally expensive. In this review, we focus on not only how machine learning is employed to speed up such excited-state simulations but also how this branch of artificial intelligence can be used to advance this exciting research field in all its aspects. Discussed applications of machine learning for excited states include excited-state dynamics simulations, static calculations of absorption spectra, as well as many others. In order to put these studies into context, we discuss the promises and pitfalls of the involved machine learning techniques. Since the latter are mostly based on quantum chemistry calculations, we also provide a short introduction into excited-state electronic structure methods and approaches for nonadiabatic dynamics simulations and describe tricks and problems when using them in machine learning for excited states of molecules.

show abstract

Section: Quantum Chemical Theory and Methodsmentioning

confidence: 99%

Section: Data Sets For Excited Statesmentioning

confidence: 99%

See 1 more Smart Citation

Machine Learning for Electronically Excited States of Molecules

2020

View full text Add to dashboard Cite

show abstract

“…Gathering data from ab initio MD prevents the latter issue, but at a higher computational cost. Some works avoid performing dynamic simulations, but still require forward exploration of the PES to find new training points 39 . Even NN simulations need to sample very large amounts of low uncertainty phase space before stumbling upon uncertain regions.…”

Section: Introductionmentioning

confidence: 99%

Differentiable sampling of molecular geometries with uncertainty-based adversarial attacks

2021

View full text Add to dashboard Cite

Neural network (NN) interatomic potentials provide fast prediction of potential energy surfaces, closely matching the accuracy of the electronic structure methods used to produce the training data. However, NN predictions are only reliable within well-learned training domains, and show volatile behavior when extrapolating. Uncertainty quantification methods can flag atomic configurations for which prediction confidence is low, but arriving at such uncertain regions requires expensive sampling of the NN phase space, often using atomistic simulations. Here, we exploit automatic differentiation to drive atomistic systems towards high-likelihood, high-uncertainty configurations without the need for molecular dynamics simulations. By performing adversarial attacks on an uncertainty metric, informative geometries that expand the training domain of NNs are sampled. When combined with an active learning loop, this approach bootstraps and improves NN potentials while decreasing the number of calls to the ground truth method. This efficiency is demonstrated on sampling of kinetic barriers, collective variables in molecules, and supramolecular chemistry in zeolite-molecule interactions, and can be extended to any NN potential architecture and materials system.

show abstract

“…Conversely, if the training set already has similar structures of the given one, the predicted results of these models should be consistent. This algorithm was also called "active learning" and has been used by many works [42][43][44][45] . Recently, a concurrent learning algorithm was proposed by E and co-workers 46 .…”

Section: Introductionmentioning

confidence: 99%

Automated Construction of Neural Network Potential Energy Surface: The Enhanced Self-Organizing Incremental Neural Network Deep Potential Method

Xu¹,

Zhu²,

Zhang³

2021

Preprint

View full text Add to dashboard Cite

In recent years, the use of deep learning (neural network) potential energy surface (NNPES) in molecular dynamics simulation has experienced explosive growth as it can be as accurate as quantum chemistry methods while being as efficient as classical mechanic methods. However, the development of NNPES is highly non-trivial. In particular, it has been troubling to construct a dataset that is as small as possible yet can cover the target chemical space. In this work, an ESOINN-DP method is developed, which has the enhanced self-organizing incremental neural-network (ESOINN) and a newly proposed error indicator at its core. With ESOINN-DP, One can construct the NNPES with little human intervention, and this method ensures that the constructed reference dataset covers the target chemical space with minimum redundancy. The performance of the ESOINN-DP method has been well validated by developing neural network potential energy surfaces for water clusters and by de-redundancy of a sub-data set of the ANI-1 database. We believe that the ESOINN-DP method provides a novelty idea for the construction of NNPES and especially, the reference datasets, and it can be used for MD simulations of various gas-phase and condensed-phase chemical systems.

show abstract

Automatically growing global reactive neural network potential energy surfaces: A trajectory-free active learning strategy

Cited by 54 publications

References 56 publications

Machine Learning for Electronically Excited States of Molecules

Machine Learning for Electronically Excited States of Molecules

Differentiable sampling of molecular geometries with uncertainty-based adversarial attacks

Automated Construction of Neural Network Potential Energy Surface: The Enhanced Self-Organizing Incremental Neural Network Deep Potential Method

Contact Info

Product

Resources

About