Extrapolation of Machine-Learning Interatomic Potentials for Organic and Polymeric Systems

Machine learning (ML) interatomic potentials (ML-IAPs) are generated for alkane and polyene hydrocarbons using on-the-fly adaptive sampling and a sparse Gaussian process regression (SGPR) algorithm. The ML model is generated based on the PBE+D3 level of density functional theory (DFT) with molecular dynamics (MD) for small alkane and polyene molecules. Intermolecular interactions are also trained with clusters and condensed phases of small molecules. It shows excellent transferability to long alkanes and closely describes the ab inito potential energy surface for polyenes. Simulation of liquid ethane also shows reasonable agreement with experimental reports. This is a promising initiative toward a universal ab initio quality force-field for hydrocarbons and organic molecules.

Recent developments in machine learning interatomic potentials (MLIPs) have empowered even nonexperts in machine learning to train MLIPs for accelerating materials simulations. However, reproducibility and independent evaluation of presented MLIP results is hindered by a lack of clear standards in current literature. In this Perspective, we aim to provide guidance on best practices for documenting MLIP use while walking the reader through the development and deployment of MLIPs including hardware and software requirements, generating training data, training models, validating predictions, and MLIP inference. We also suggest useful plotting practices and analyses to validate and boost confidence in the deployed models. Finally, we provide a step-by-step checklist for practitioners to use directly before publication to standardize the information to be reported. Altogether, we hope that our work will encourage the reliable and reproducible use of these MLIPs, which will accelerate their ability to make a positive impact in various disciplines including materials science, chemistry, and biology, among others.

In nuclear and/or solar applications that involve molten salts, impurities frequently enter the salt as either fission products or via corrosion. Impurities can interact and make complexes, but the impact of such complexation on the properties of the salts and corrosion rates has not been understood. Common impurities in molten salts, such as FLiBe, include Cr, Ni, and Fe. Here, we investigate the complexation of Cr and Ni in FLiBe using molecular dynamics based on a machine learning interatomic potential (MLIP) fitted using the atomic cluster expansion (ACE) method. The MLIP allows us to overcome the challenges of simultaneously needing accurate energetics and long time scale to study complexation. We demonstrate that impurity behavior is more difficult to capture than that of concentrated elements with MLIPs due to less sampling in training data, but that this can be overcome by using active learning strategies to obtain a robust fit. Our findings suggest that there is a weak but potentially significant binding free energy between CrF2 and NiF2 in eutectic FLiBe of -0.112eV. Under certain conditions this binding creates a significant short-range order between the species and lowers the redox potential of NiF2 in the presence of CrF2 in FLiBe, making Ni dissolution more favorable in the presence of Cr as compared to its dissolution in pure FLiBe. However, we find little impact of this complexation on the diffusivity of Ni and Cr. Overall, the methodology presented here suggests an approach to modeling complexation with MLIPs and suggests that interactions between dissolved cations could be playing a significant role in some salt thermophysical properties.

Machine-learning interatomic potentials (MLIPs) have become a mainstay in computationally guided materials science, surpassing traditional force fields due to their flexible functional form and superior accuracy in reproducing physical properties of materials. This flexibility is achieved through mathematically rigorous basis sets that describe interatomic interactions within a local atomic environment. The number of parameters in these basis sets influences both the size of the training dataset required and the computational speed of the MLIP. Consequently, compressing MLIPs by reducing the number of parameters is a promising route to more efficient simulations. In this work, we use low-rank matrix and tensor factorizations under fixed-rank constraints to achieve this compression. In addition, we demonstrate that an algorithm with automatic rank augmentation helps to find a deeper local minimum of the fitted potential. The methodology is mainly verified using the Moment Tensor Potential (MTP) model and benchmarked on multi-component systems: a Mo-Nb-Ta-W medium-entropy alloy, molten LiF-NaF-KF, and a glycine molecular crystal. The proposed approach achieves up to 50% compression without any loss of MTP accuracy. We also demonstrate that the developed methodology is universal and can be applied to compress other MLIPs on the example of atomic cluster expansion.

Machine learning interatomic potentials (MLIPs) enable more efficient molecular dynamics (MD) simulations with ab initio accuracy, which have been used in various domains of physical science. However, distribution shift between training and test data causes deterioration of the test performance of MLIPs, and even leads to collapse of MD simulations. In this work, we propose an online Test-time Adaptation Interatomic Potential (TAIP) framework to improve the generalization on test data. Specifically, we design a dual-level self-supervised learning approach that leverages global structure and atomic local environment information to align the model with the test data. Extensive experiments demonstrate TAIP’s capability to bridge the domain gap between training and test dataset without additional data. TAIP enhances the test performance on various benchmarks, from small molecule datasets to complex periodic molecular systems with various types of elements. TAIP also enables stable MD simulations where the corresponding baseline models collapse.

We present an algorithm for accelerating the search of a molecule's adsorption sites based on global optimization of surface adsorbate geometries. Our approach uses a machine-learning interatomic potential (moment tensor potential) to approximate the potential energy surface and an active learning algorithm for the automatic construction of an optimal training dataset. To validate our methodology, we compare the results across various well-known catalytic systems with surfaces of different crystallographic orientations and adsorbate geometries, including CO/Pd(111), NO/Pd(100), NH3/Cu(100), C6H6/Ag(111), and CH2CO/Rh(211). In all the cases, we observed an agreement of our results with the literature.

The spontaneous internalization of electrostatic interactions by large atomic models in molten salts

Using machine learning (ML) to construct interatomic interactions and thus potential energy surface (PES) has become a common strategy for materials design and simulations. However, those current models of machine-learning interatomic potential (MLIP) consider no relevant physical constraints or global scaling and thus may owe intrinsic out-of-domain difficulty which underlies the challenges of model generalizability and physical scalability. Here, by incorporating the global universal scaling law, we develop an ultrasmall parameterized MLIP with superlinear expressive capability, named SUS2-MLIP. Due to the global scaling derived from the universal equation of state (UEOS), SUS2-MLIP not only has significantly reduced parameters by decoupling the element space from coordinate space but also naturally outcomes the out-of-domain difficulty and endows the model with inherent generalizability and scalability even with relatively small training dataset. The non-linearity-embedding transformation in radial function endows the model with superlinear expressive capability. SUS2-MLIP outperforms the state-of-the-art MLIP models with its exceptional computational efficiency, especially for multiple-element materials and physical scalability in property prediction. This work not only presents a highly efficient universal MLIP model but also sheds light on incorporating physical constraints into AI-aided materials simulation.

Machine learning interatomic potentials (MLIPs) enables molecular dynamics (MD) simulations with ab initio accuracy and has been applied to various fields of physical science. However, the performance and transferability of MLIPs are limited by insufficient labeled training data, which require expensive ab initio calculations to obtain the labels, especially for complex molecular systems. To address this challenge, we design a novel geometric structure learning paradigm that consists of two stages. We first generate a large quantity of 3D configurations of target molecular system with classical molecular dynamics simulations. Then, we propose geometry-enhanced self-supervised learning consisting of masking, denoising, and contrastive learning to better capture the topology and 3D 1 geometric information from the unlabeled 3D configurations. We evaluate our method on various benchmarks ranging from small molecule datasets to complex periodic molecular systems with more types of elements. The experimental results show that the proposed pre-training method can greatly enhance the accuracy of MLIPs with few extra computational costs and works well with different invariant or equivariant graph neural network architectures. Our method improves the generalization capability of MLIPs and helps to realize accurate MD simulations for complex molecular systems.

Adaptive loss weighting for machine learning interatomic potentials

Existing SAR ATR systems are usually trained off-line with samples of target imagery or CAD models, prior to conducting a mission. If the training data is not representative of mission conditions, then poor performance may result. In addition, it is difficult to acquire suitable training data for the many target types of interest. The Adaptive SAR ATR Problem Set (AdaptSAPS) program provides a MATLAB framework and image database for developing systems that adapt to mission conditions, meaning less reliance on accurate training data. A key function of an adaptive system is the ability to utilise truth feedback to improve performance, and it is this feature which AdaptSAPS is intended to exploit. This paper presents a new method for SAR ATR that does not use training data, based on supervised learning. This is achieved by using feature-based classification, and several new shadow features have been developed for this purpose. These features allow discrimination of vehicles from clutter, and classification of vehicles into two classes: targets, comprising military combat types, and non-targets, comprising bulldozers and trucks. The performance of the system is assessed using three baseline missions provided with AdaptSAPS, as well as three additional missions. All performance metrics indicate a distinct learning trend over the course of a mission, with most third and fourth quartile performance levels exceeding 85% correct classification. It has been demonstrated that these performance levels can be maintained even when truth feedback rates are reduced by up to 55% over the course of a mission.

Phase unwrapping is an essential process in synthetic aperture radar interferometry that restores phase signals constrained within the range of (-π, π) to their true phase values. Traditional algorithm-based methods can introduce significant errors due to rapid and steep phase gradient or noise, which negatively impact terrain elevation and surface displacement analyses. To overcome these limitations, deep learningbased phase unwrapping techniques have been proposed; however, there has been insufficient previous studies due to the lack of accurate training and test data. This paper aims to share the training data generated using the phase unwrapping simulation method with locally-different phase noise. The data were generated by the simulation of topographic phases and phase noise, atmospheric and orbital distortions. Additionally, data augmentation for phase variation and noise levels was applied to address data imbalance issues. The shared data consists of two types: one with a constant phase noise level for each patch, and another that simulates locally different phase noise based on augmented coherence data. This data is primarily effective for unwrapping topographic phase components and holds significance as the first phase unwrapping training data of synthetic aperture radar interferograms shared in Korea. We expect this resource to serve as foundational data for future phase unwrapping technology research, including applications for upcoming satellites like KOMPSAT-6 and water resource satellites.

A critical part in preoperative planning for revision arthroplasty surgery involves the identification of the failed implant. Using a predictive artificial neural network (ANN) model, the objectives of this study were: (1) to develop a machine-learning algorithm using operative big data to identify an implant from a radiograph; and (2) to compare algorithms that optimise accuracy in a timely fashion. Using 2116 postoperative anteroposterior (AP) hip radiographs of total hip arthroplasties from 2002 to 2019, 10 artificial neural networks were modeled and trained to classify the radiograph according to the femoral stem implanted. Stem brand and model was confirmed with 1594 operative reports. Model performance was determined by classification accuracy toward a random 706 AP hip radiographs, and again on a consecutive series of 324 radiographs prospectively collected over 2019. The Dense-Net 201 architecture outperformed all others with 100.00% accuracy in training data, 95.15% accuracy on validation data, and 91.16% accuracy in the unique prospective series of patients. This outperformed all other models on the validation (p < 0.0001) and novel series (p < 0.0001). The convolutional neural network also displayed the probability (confidence) of the femoral stem classification for any input radiograph. This neural network averaged a runtime of 0.96 (SD 0.02) seconds for an iPhone 6 to calculate from a given radiograph when converted to an application. Neural networks offer a useful adjunct to the surgeon in preoperative identification of the prior implant.

Financial institutions have a large amount of data on their borrowers, which can be used to predict the probability of borrowers defaulting their loan or not. Some of the models that have been used to predict individual loan defaults include linear discriminant analysis models and extreme value theory models. These models are parametric in nature since they assume that the response being investigated takes a particular functional form. However, there is a possibility that the functional form used to estimate the response is very different from the actual functional form of the response. The purpose of this research was to analyze individual loan defaults in Kenya using the logistic regression model. The data used in this study was obtained from equity bank of Kenya for the period between 2006 to 2016. A random sample of 1000 loan applicants whose loans had been approved by equity bank of Kenya during this period was obtained. Data obtained was on the credit history, purpose of the loan, loan amount, nature of the saving account, employment status, sex of the applicant, age of the applicant, security used when acquiring the loan and the area of residence of the applicant (rural or urban). This study employed a quantitative research design, it deals with individual loans defaults as group characteristics of a borrower. The data was pre-processed by seeding using R- Software and then split into training dataset and test data set. The train data was used to train the logistic regression model by employing Supervised machine learning approach. The R-statistical software was used for the analysis of the data. The test data set was used to do cross-validation of the developed logistic model which later was used for analysis prediction of individual loan defaults. This study focused on the analysis of individual loan defaults in Kenya using the logistic regression model in Machine learning. The logistic regression model predicted 303 defaults from train data set, 122 non-defaults and misclassified loans were 56 and 69. The model had an accuracy of 0.7727 with the train data and 0.7333 with the test data. The logistic regression model showed a precision of 0.8440 and 0.8244 with the train and test data respectively. The performance of the model with both the train and test data was illustrated using a plot of train errors and test errors against sample size on the same axes. The plot showed that the performance of the model increases with an increase in sample size. The study recommended the use of logistic regression in conjunction with supervised machine learning approach in loan default prediction in financial institutions and also more research should be carried out on ensemble methods of loan defaults prediction in order to increase the prediction accuracy.

In this work, we present enhanced representation-based sampling (ERBS), a novel enhanced sampling method designed to generate structurally diverse training data sets for machine-learned interatomic potentials. ERBS automatically identifies collective variables by dimensionality reduction of atomic descriptors and applies a bias potential inspired by the On-the-Fly probability enhanced sampling framework. We highlight the ability of Gaussian moment descriptors to capture collective molecular motions and explore the impact of biasing parameters using alanine dipeptide as a benchmark system. We show that free energy surfaces can be reconstructed with high fidelity using only short biased trajectories as training data. Further, we apply the method to the iterative construction of a liquid water data set and compare the quality of simulated self-diffusion coefficients for models trained with molecular dynamics and ERBS data. Further, we active-learn models for liquid water with and without enhanced sampling and compare the quality of simulated self-diffusion coefficients. The self-diffusion coefficients closely match those simulated with a reference model at a significantly reduced data set size. Finally, we compare the sampling behavior of enhanced sampling methods by benchmarking the mean squared displacements of BMIM+BF4- trajectories simulated with uncertainty-driven dynamics and ERBS and find that the latter significantly increases the exploration of configurational space.