scholarly journals Active learning of reactive Bayesian force fields: Application to heterogeneous catalysis dynamics of H/Pt

2021 ◽  
Author(s):  
Jonathan Vandermause ◽  
Yu Xie ◽  
Jin Soo Lim ◽  
Cameron Owen ◽  
Boris Kozinsky
Keyword(s):  
Active Learning ◽  
Ab Initio ◽  
Force Fields ◽  
Reactive Systems ◽  
Training Data ◽  
Many Body ◽  
Time Step ◽  
Two Phase ◽  
Learning Framework ◽  
Dynamics Simulations

Abstract Atomistic modeling of chemically reactive systems has so far relied on either expensive ab initio methods or bond-order force fields requiring arduous parametrization. Here, we describe a Bayesian active learning framework for autonomous ``on-the-fly'' training of fast and accurate reactive many-body force fields during molecular dynamics simulations. At each time step, predictive uncertainties of a sparse Gaussian process are evaluated to automatically determine whether additional ab initio training data are needed. We introduce a general method for mapping trained kernel models onto equivalent polynomial models whose prediction cost is much lower and independent of the training set size. As a demonstration, we perform direct two-phase simulations of heterogeneous H2 turnover on the Pt(111) catalyst surface, at chemical accuracy. The model trains itself in three days and performs at twice the speed of a ReaxFF model, while maintaining much higher fidelity to DFT and excellent agreement with experiment.

scholarly journals A universal framework for featurization of atomistic systems

2021 ◽  
Author(s):  
Xiangyun Lei ◽  
Andrew Medford
Keyword(s):  
Machine Learning ◽  
Molecular Dynamics ◽  
Force Fields ◽  
Reactive Systems ◽  
Quantum Molecular Dynamics ◽  
Learning Models ◽  
Fixed Dimension ◽  
Comparable Performance ◽  
Invaluable Tool ◽  
Dynamics Simulations

Abstract Molecular dynamics simulations are an invaluable tool in numerous scientific fields. However, the ubiquitous classical force fields cannot describe reactive systems, and quantum molecular dynamics are too computationally demanding to treat large systems or long timescales. Reactive force fields based on physics or machine learning can be used to bridge the gap in time and length scales, but these force fields require substantial effort to construct and are highly specific to a given chemical composition and application. A significant limitation of machine learning models is the use of element-specific features, leading to models that scale poorly with the number of elements. This work introduces the Gaussian multipole (GMP) featurization scheme that utilizes physically-relevant multipole expansions of the electron density around atoms to yield feature vectors that interpolate between element types and have a fixed dimension regardless of the number of elements present. We combine GMP with neural networks to directly compare it to the widely used Behler-Parinello symmetry functions for the MD17 dataset, revealing that it exhibits improved accuracy and computational efficiency. Further, we demonstrate that GMP-based models can achieve chemical accuracy for the QM9 dataset, and their accuracy remains reasonable even when extrapolating to new elements. Finally, we test GMP-based models for the Open Catalysis Project (OCP) dataset, revealing comparable performance to graph convolutional deep learning models. The results indicate that this featurization scheme fills a critical gap in the construction of efficient and transferable machine-learned force fields.

Empirical Potential-Energy Surfaces Fitting Using Feed forward Neural Networks

2012 ◽  
Author(s):  
Lionel Raff ◽  
Ranga Komanduri ◽  
Martin Hagan ◽  
Satish Bukkapatnam
Keyword(s):  
Potential Energy ◽  
Ab Initio ◽  
Potential Energy Surfaces ◽  
Force Fields ◽  
Pair Potential ◽  
Potential Surfaces ◽  
Empirical Potential ◽  
Functional Forms ◽  
Dynamics Simulations ◽  
Energy Surfaces

When the system of interest becomes too complex to permit the use of ab initio methods to obtain the system potential-energy surfaces (PES), empirical potential surfaces are frequently employed to represent the force fields present in the system under investigation. In most cases, the functional forms present in these potentials are selected on the basis of chemical and physical intuitions. The parameters of the surface are frequently adjusted to fit a very small set of experimental data that comprise bond energies, equilibrium bond distances and angles, fundamental vibrational frequencies, and perhaps measured barrier heights to reactions of interest. Such potentials generally yield only qualitative or semiquantitative descriptions of the system dynamics. Several research groups have significantly improved the accuracy of the values of the experimental properties computed using empirical potential surfaces by fitting the chosen functional form for the potential to the force fields obtained from trajectories using ab initio Car-Parrinello molecular dynamics simulations. The fitting to the force fields is usually done using a least-squares fitting approach. This method has been employed by Izvekov et al. to obtain effective non-polarizable three-site force fields for liquid water. Carré et al. have employed such a procedure to obtain a new pair potential for silica. In their investigation, the vector of potential parameters was fitted using an iterative Levenberg-Marquardt algorithm. Tangney and Scandolo have also developed an interatomic force field for liquid SiO2 in which the parameters were fitted to the forces, stresses, and energies obtained from ab initio calculations. Ercolessi and Adams have used a quasi-Newtonian procedure to fit an empirical potential for aluminum to data obtained from first-principals computations. Empirical potentials can be improved by making the parameters parameterized functions of the coordinates defining the instantaneous positions of the atoms of the system. This approach has been successfully employed by numerous investigators The difficulty with this procedure is that the number of parameters that must be adjusted increases rapidly. Appropriate fitting of these parameters requires a much more extensive database. Finally, the actual fitting process can often be tedious, difficult, and time-consuming.

The Future

1992 ◽  
Author(s):  
John A. Tossell ◽  
David J. Vaughan
Keyword(s):  
Ab Initio ◽  
Quantum Chemical ◽  
Force Fields ◽  
Structural Characteristics ◽  
Future Research ◽  
Quantum Mechanical ◽  
Bond Lengths ◽  
Hartree Fock ◽  
Qualitative Level ◽  
Dynamics Simulations

In this final chapter, an attempt is made to provide an overview of the capabilities of quantum-mechanical methods at the present time, and to highlight the needs for future development and possible future applications of these methods, particularly in areas related to mineral structures, energetics, and spectroscopy. There is also a brief account of some new areas of application, specific directions for future research, and possible developments in the perception and use of quantum-mechanical approaches. The book ends with an epilog on the overall role of “theoretical geochemistry” in the earth and environmental sciences. The local structural characteristics of minerals such as Mg2SiO4, which contain only main-group elements, are reasonably well reproduced by ab initio Hartree-Fock-Roothaan (SCF) cluster calculations at the mediumbasis- set level. Calculations incorporating configuration interaction will inevitably follow and probably lead to somewhat better agreement with experiment. The most pressing needs in this area of study are for the development of systematic procedures for cluster selection and embedding, for a greater understanding of the results at a qualitative level, and for more widespread efficient application of the quantum-chemical results currently available. In the last area, substantial progress has already been made by Lasaga and Gibbs (1987), Sanders et al. (1984), Tsuneyuki et al. (1988), and others, who have used ab initio calculations to generate theoretical force fields which can then be used in molecular-dynamics simulations. If the characteristics of the resultant force fields can be understood at a first-principles level, then it may be possible to understand details of the simulated structures at the same level. Unfortunately, as regards a greater qualitative understanding of the quantum-mechanical calculations, little progress has been made. Rather old qualitative theories describe some aspects of bond-angle variation (Tossell, 1986), but no general model to interpret variations in bond lengths has been developed within either chemistry or geochemistry beyond the model of additive atomic (Slater) or ionic (Shannon and Prewitt) radii. Indeed, global theories of bond-length variations within an ab initio framework seem to be nonexistent. Nonetheless, quantum-chemical studies have shown the presence of intriguing systematics in bond lengths (Gibbs et al., 1987), which had been already noted empirically.

New Technique for AB Initio Atomistic Potentials and Application to Thermal Expansion of Ni/Cr Alloys

1992 ◽  
Vol 291 ◽  
Author(s):  
J. Mei ◽  
B.R. Cooper ◽  
Y.G. Hao ◽  
S.P. Lim ◽  
F.L. VanScoy
Keyword(s):  
Thermal Expansion ◽  
Ab Initio ◽  
Density Functional ◽  
Interatomic Potential ◽  
Coefficient Of Thermal Expansion ◽  
Many Body ◽  
Functional Theory ◽  
Transition Metal Alloys ◽  
Wide Range ◽  
Dynamics Simulations

ABSTRACTA scheme of developing ab initio many body potentials based on total energy calculations within density functional theory (DFT) is presented and demonstrated for transition metal alloys. An ab initio interatomic potential for Ni/Cr alloys is constructed with no input from experimental data. Molecular dynamics simulations have been performed to study thermal expansions. The coefficient of thermal expansion (CTE) has been calculated over a wide range of temperature, and good agreement is obtained between theory and experiment.

A Semi-Supervised Active Learning Framework for Image Classification

2014 ◽  
Vol 556-562 ◽  
pp. 4765-4769
Author(s):  
Han Yi Li ◽  
Ming Yang ◽  
Nan Nan Kang ◽  
Lu Lu Yue
Keyword(s):  
Active Learning ◽  
Image Classification ◽  
Supervised Learning ◽  
The Other ◽  
Training Data ◽  
Classification Error ◽  
Classification Algorithms ◽  
Learning Framework ◽  
User Query ◽  
Changing Rate

In this paper, a novel image classification method, incorporating active learning and semi-supervised learning (SSL), is proposed. In this method, two classifiers are needed where one is trained by labeled data and some unlabeled data, while the other one is trained only by labeled data. Specifically, in each round, two classifiers iterate to select useful examples in contention for user query. Then we compute the label changing rate for every unlabeled example in each classifier. Those examples in which the label changing rate is zero and the label in the two classifiers is the same are selected to add into the training data of the first classifier. Our experimental results show that our method significantly reduced the need of labeled examples, while at the same time reducing classification error compared with widely used image classification algorithms.

Constructing ab initio force fields for molecular dynamics simulations

1998 ◽  
Vol 108 (12) ◽  
pp. 4739-4755 ◽  
Author(s):  
Yi-Ping Liu ◽  
Kyungsun Kim ◽  
B. J. Berne ◽  
Richard A. Friesner ◽  
Steven W. Rick
Keyword(s):  
Molecular Dynamics ◽  
Ab Initio ◽  
Molecular Dynamics Simulations ◽  
Force Fields ◽  
Dynamics Simulations

scholarly journals A Transfer Learning–Based Active Learning Framework for Brain Tumor Classification

2021 ◽  
Vol 4 ◽  
Author(s):  
Ruqian Hao ◽  
Khashayar Namdar ◽  
Lin Liu ◽  
Farzad Khalvati
Keyword(s):  
Deep Learning ◽  
Brain Tumor ◽  
Active Learning ◽  
Transfer Learning ◽  
Medical Images ◽  
Tumor Classification ◽  
Training Data ◽  
Validation Dataset ◽  
Low Grade ◽  
Learning Framework

Brain tumor is one of the leading causes of cancer-related death globally among children and adults. Precise classification of brain tumor grade (low-grade and high-grade glioma) at an early stage plays a key role in successful prognosis and treatment planning. With recent advances in deep learning, artificial intelligence–enabled brain tumor grading systems can assist radiologists in the interpretation of medical images within seconds. The performance of deep learning techniques is, however, highly depended on the size of the annotated dataset. It is extremely challenging to label a large quantity of medical images, given the complexity and volume of medical data. In this work, we propose a novel transfer learning–based active learning framework to reduce the annotation cost while maintaining stability and robustness of the model performance for brain tumor classification. In this retrospective research, we employed a 2D slice–based approach to train and fine-tune our model on the magnetic resonance imaging (MRI) training dataset of 203 patients and a validation dataset of 66 patients which was used as the baseline. With our proposed method, the model achieved area under receiver operating characteristic (ROC) curve (AUC) of 82.89% on a separate test dataset of 66 patients, which was 2.92% higher than the baseline AUC while saving at least 40% of labeling cost. In order to further examine the robustness of our method, we created a balanced dataset, which underwent the same procedure. The model achieved AUC of 82% compared with AUC of 78.48% for the baseline, which reassures the robustness and stability of our proposed transfer learning augmented with active learning framework while significantly reducing the size of training data.

Sign in / Sign up

Export Citation Format

Share Document