Search

The Journal of Chemical Physics has released a special issue focused on software for atomistic machine learning. This issue aims to address the lack of journals dedicated to publishing scientific software papers. The collection of papers in this issue provides insight into the tools and goals of software implementations in the field of atomistic machine learning. The articles cover a range of topics, including machine-learning interatomic potentials, sampling, dataset repositories, workflows, and auxiliary tooling and analysis. The article concludes by emphasizing the importance of software implementations in the field and encourages further submissions on relevant topics. [Extracted from the article]

Data generation remains a bottleneck in training surrogate models to predict molecular properties. We demonstrate that multitask Gaussian process regression overcomes this limitation by leveraging both expensive and cheap data sources. In particular, we consider training sets constructed from coupled-cluster (CC) and density functional theory (DFT) data. We report that multitask surrogates can predict at CC-level accuracy with a reduction in data generation cost by over an order of magnitude. Of note, our approach allows the training set to include DFT data generated by a heterogeneous mix of exchange–correlation functionals without imposing any artificial hierarchy on functional accuracy. More generally, the multitask framework can accommodate a wider range of training set structures—including the full disparity between the different levels of fidelity—than existing kernel approaches based on Δ-learning although we show that the accuracy of the two approaches can be similar. Consequently, multitask regression can be a tool for reducing data generation costs even further by opportunistically exploiting existing data sources. [ABSTRACT FROM AUTHOR]

We investigate the endofullerene system 3He@C60 with a four-dimensional potential energy surface (PES) to include the three He translational degrees of freedom and C60 cage radius. We compare second order Møller–Plesset perturbation theory (MP2), spin component scaled-MP2, scaled opposite spin-MP2, random phase approximation (RPA)@Perdew, Burke, and Ernzerhof (PBE), and corrected Hartree–Fock-RPA to calibrate and gain confidence in the choice of electronic structure method. Due to the high cost of these calculations, the PES is interpolated using Gaussian Process Regression (GPR), owing to its effectiveness with sparse training data. The PES is split into a two-dimensional radial surface, to which corrections are applied to achieve an overall four-dimensional surface. The nuclear Hamiltonian is diagonalized to generate the in-cage translational/vibrational eigenstates. The degeneracy of the three-dimensional harmonic oscillator energies with principal quantum number n is lifted due to the anharmonicity in the radial potential. The (2l + 1)-fold degeneracy of the angular momentum states is also weakly lifted, due to the angular dependence in the potential. We calculate the fundamental frequency to range between 96 and 110 cm−1 depending on the electronic structure method used. Error bars of the eigenstate energies were calculated from the GPR and are on the order of ∼±1.5 cm−1. Wavefunctions are also compared by considering their overlap and Hellinger distance to the one-dimensional empirical potential. As with the energies, the two ab initio methods MP2 and RPA@PBE show the best agreement. While MP2 has better agreement than RPA@PBE, due to its higher computational efficiency and comparable performance, we recommend RPA as an alternative electronic structure method of choice to MP2 for these systems. [ABSTRACT FROM AUTHOR]

A merged potential energy surface (PES) is introduced for CO + CO collisions by combining a recent full-dimensional ab initio PES [Chen et al. J. Chem. Phys. 153, 054310 (2020)] and analytical long-range multipolar interactions. This merged PES offers a double advantage: it retains the precision of the ab initio PES in describing the van der Waals well and repulsive short range while providing an accurate physical description of long-range interaction; it significantly reduces the computational time required for trajectory integration since the long-range portion of the ab initio PES (involving numerous neural network fitting parameters) is now replaced by the analytical model potential. Based on the present merged PES, mixed Quantum-Classical (MQC) calculations, which capture quantum effects related to vibrational motion, align with a range of experimental data, including transport properties, vibrational energy transfer between CO and its isotoplogues, as well as rate coefficients for V–V and V–T/R processes. Notably, the original ab initio PES yields V–T/R rate coefficients at low temperatures that are significantly higher than the experimental data due to the artificial contribution of its unphysical long-range potential. In addition to conducting extensive MQC calculations to obtain raw data for V–V and V–T/R rate coefficients, we employ Gaussian process regression to predict processes lacking computed MQC data, thereby completing the considered V–V and V–T/R datasets. These extensive rate coefficient datasets, particularly for V–T/R processes, are unprecedented and reveal the significant role played by V–T/R processes at high temperatures, emphasizing the necessity of incorporating both V–V and V–T/R processes in the applications. [ABSTRACT FROM AUTHOR]

Kernel methods such as kernel ridge regression and Gaussian process regression with Matern-type kernels have been increasingly used, in particular, to fit potential energy surfaces (PES) and density functionals, and for materials informatics. When the dimensionality of the feature space is high, these methods are used with necessarily sparse data. In this regime, the optimal length parameter of a Matern-type kernel may become so large that the method effectively degenerates into a low-order polynomial regression and, therefore, loses any advantage over such regression. This is demonstrated theoretically as well as numerically in the examples of six- and fifteen-dimensional molecular PES using squared exponential and simple exponential kernels. The results shed additional light on the success of polynomial approximations such as PIP for medium-size molecules and on the importance of orders-of-coupling-based models for preserving the advantages of kernel methods with Matern-type kernels of on the use of physically motivated (reproducing) kernels. [ABSTRACT FROM AUTHOR]

Machine learning (ML) of kinetic energy functionals (KEFs), in particular kinetic energy density (KED) functionals, is a promising way to construct KEFs for orbital-free density functional theory (DFT). Neural networks and kernel methods including Gaussian process regression (GPR) have been used to learn Kohn–Sham (KS) KED from density-based descriptors derived from KS DFT calculations. The descriptors are typically expressed as functions of different powers and derivatives of the electron density. This can generate large and extremely unevenly distributed datasets, which complicates effective application of ML techniques. Very uneven data distributions require many training datapoints, can cause overfitting, and can ultimately lower the quality of an ML KED model. We show that one can produce more accurate ML models from fewer data by working with smoothed density-dependent variables and KED. Smoothing palliates the issue of very uneven data distributions and associated difficulties of sampling while retaining enough spatial structure necessary for working within the paradigm of KEDF. We use GPR as a function of smoothed terms of the fourth order gradient expansion and KS effective potential and obtain accurate and stable (with respect to different random choices of training points) kinetic energy models for Al, Mg, and Si simultaneously from as few as 2000 samples (about 0.3% of the total KS DFT data). In particular, accuracies on the order of 1% in a measure of the quality of energy–volume dependence B ′ = E V 0 − Δ V − 2 E V 0 + E ( V 0 + Δ V) Δ V / V 0 2 (where V0 is the equilibrium volume and ΔV is a deviation from it) are obtained simultaneously for all three materials. [ABSTRACT FROM AUTHOR]

Free energy simulations that employ combined quantum mechanical and molecular mechanical (QM/MM) potentials at ab initio QM (AI) levels are computationally highly demanding. Here, we present a machine-learning-facilitated approach for obtaining AI/MM-quality free energy profiles at the cost of efficient semiempirical QM/MM (SE/MM) methods. Specifically, we use Gaussian process regression (GPR) to learn the potential energy corrections needed for an SE/MM level to match an AI/MM target along the minimum free energy path (MFEP). Force modification using gradients of the GPR potential allows us to improve configurational sampling and update the MFEP. To adaptively train our model, we further employ the sparse variational GP (SVGP) and streaming sparse GPR (SSGPR) methods, which efficiently incorporate previous sample information without significantly increasing the training data size. We applied the QM-(SS)GPR/MM method to the solution-phase SN2 Menshutkin reaction, N H 3 + C H 3 C l → C H 3 N H 3 + + C l − , using AM1/MM and B3LYP/6-31+G(d,p)/MM as the base and target levels, respectively. For 4000 configurations sampled along the MFEP, the iteratively optimized AM1-SSGPR-4/MM model reduces the energy error in AM1/MM from 18.2 to 4.4 kcal/mol. Although not explicitly fitting forces, our method also reduces the key internal force errors from 25.5 to 11.1 kcal/mol/ A ̊ and from 30.2 to 10.3 kcal/mol/ A ̊ for the N–C and C–Cl bonds, respectively. Compared to the uncorrected simulations, the AM1-SSGPR-4/MM method lowers the predicted free energy barrier from 28.7 to 11.7 kcal/mol and decreases the reaction free energy from −12.4 to −41.9 kcal/mol, bringing these results into closer agreement with their AI/MM and experimental benchmarks. [ABSTRACT FROM AUTHOR]

We present a new program implementation of the Gaussian process regression adaptive density-guided approach [Schmitz et al., J. Chem. Phys. 153, 064105 (2020)] for automatic and cost-efficient potential energy surface construction in the MidasCpp program. A number of technical and methodological improvements made allowed us to extend this approach toward calculations of larger molecular systems than those previously accessible and maintain the very high accuracy of constructed potential energy surfaces. On the methodological side, improvements were made by using a Δ-learning approach, predicting the difference against a fully harmonic potential, and employing a computationally more efficient hyperparameter optimization procedure. We demonstrate the performance of this method on a test set of molecules of growing size and show that up to 80% of single point calculations could be avoided, introducing a root mean square deviation in fundamental excitations of about 3 cm−1. A much higher accuracy with errors below 1 cm−1 could be achieved with tighter convergence thresholds still reducing the number of single point computations by up to 68%. We further support our findings with a detailed analysis of wall times measured while employing different electronic structure methods. Our results demonstrate that GPR-ADGA is an effective tool, which could be applied for cost-efficient calculations of potential energy surfaces suitable for highly accurate vibrational spectra simulations. [ABSTRACT FROM AUTHOR]

The theoretical investigation of gas adsorption, storage, separation, diffusion, and related transport processes in porous materials relies on a detailed knowledge of the potential energy surface of molecules in a stationary environment. In this article, a new algorithm is presented, specifically developed for gas transport phenomena, which allows for a highly cost-effective determination of molecular potential energy surfaces. It is based on a symmetry-enhanced version of Gaussian process regression with embedded gradient information and employs an active learning strategy to keep the number of single point evaluations as low as possible. The performance of the algorithm is tested for a selection of gas sieving scenarios on porous, N-functionalized graphene and for the intermolecular interaction of CH4 and N2. [ABSTRACT FROM AUTHOR]

We present a local and transferable machine-learning approach capable of predicting the real-space density response of both molecules and periodic systems to homogeneous electric fields. The new method, Symmetry-Adapted Learning of Three-dimensional Electron Responses (SALTER), builds on the symmetry-adapted Gaussian process regression symmetry-adapted learning of three-dimensional electron densities framework. SALTER requires only a small, but necessary, modification to the descriptors used to represent the atomic environments. We present the performance of the method on isolated water molecules, bulk water, and a naphthalene crystal. Root mean square errors of the predicted density response lie at or below 10% with barely more than 100 training structures. Derived polarizability tensors and even Raman spectra further derived from these tensors show good agreement with those calculated directly from quantum mechanical methods. Therefore, SALTER shows excellent performance when predicting derived quantities, while retaining all of the information contained in the full electronic response. Thus, this method is capable of predicting vector fields in a chemical context and serves as a landmark for further developments. [ABSTRACT FROM AUTHOR]

Secondary metabolites have high ecological and economic values. Measuring the amounts of these metabolites in plants through laboratory analysis is time-consuming and expensive. On the other hand, the laboratory results cannot be generalized to a wide area of rangelands with diverse climates, topography, and soil. Therefore, the present study aims to evaluate the potential of PRISMA hyperspectral data to estimate the rangeland phenol concentration at landscape scale in western Isfahan province, Iran. Phenol concentration and spectral signatures were extracted from 42 plant individuals, including Astragalus adscendens, Astragalus verus, Daphne mucronata, Anabasis aphylla, and Phlomis olivieri using laboratory methods and spectroscopy in the range of 350 to 2500 nm. Phenol-sensitive bands were identified by establishing a relationship between phenol concentration and all spectral bands by the Partial Least Square Regression (PLSR) method. The map of phenol spatial changes was then produced with the Bayesian Regularized Artificial Neural Network Model (BRNN), Random Forest (RF), Generalized Linear Model (GAM), Generalized Additive Model (GLM), and Gaussian Process Regression (GPR) in rangelands. The results showed that the phenol concentration in different individuals was between 3.65 and 20.13 (mgGAE/gDW), and species spectral absorptions ranged from 1.650 to 1.660. The very high coefficient of determination of phenol concentration with spectral data (R2 = 0.94) confirmed the use of these data in phenol mapping at the landscape scale. The maximum estimated phenol concentration by GLM, GAM, RF, BRNN, and GPR models was 36.40, 26.76, 18.07, 22.74, and 32.92 GAE/g DW, respectively; which RF model had the highest (R2 = 0.955) and GLM had the lowest accuracy (R2 = 0.816). The findings show that hyperspectral data can estimate secondary metabolites at the landscape scale and can supplement laboratory methods in managing rangeland plants. [ABSTRACT FROM AUTHOR]

Continuously reinforced concrete pavement (CRCP), crucial for the resilience of transportation infrastructure owing to its continuous steel reinforcement, confronts a critical challenge in the form of spalling—a distress phenomenon posing a threat to pavement durability and overall structural integrity. The detachment or breakage of concrete from the surface compromises CRCP's functionality and raises safety concerns and escalating maintenance costs. To address this pressing issue, our study investigates the multifaceted factors influencing spalling, employing a comprehensive approach that integrates statistical and machine learning techniques for predictive modeling. Descriptive statistics meticulously profile the dataset, emphasizing age, thickness, precipitation, temperature, and traffic parameters. Regression analysis unveils key relationships, emphasizing the significance of age, annual temperature, annual precipitation, maximum humidity, and the initial International Roughness Index (IRI) as influential factors. The correlation matrix heatmap guides feature selection, elucidating intricate interdependencies. Simultaneously, feature importance analysis identifies age, Average Annual Daily Traffic (AADT), and total pavement thickness as crucial contributors to spalling. In machine learning, adopting models, including Gaussian Process Regression and ensemble tree models, is grounded in their diverse capabilities and suitability for the complex task at hand. Their varying predictive accuracies underscore the importance of judicious model selection. This research advances pavement engineering practices by offering nuanced insights into factors influencing spalling in CRCP, refining our understanding of spalling influences. Consequently, the study not only opens avenues for developing improved predictive methodologies but also enhances the durability of CRCP infrastructure, addressing broader implications for informed decision-making in transportation infrastructure management. The selection of Gaussian Process Regression and ensemble tree models stems from their adaptability to capture intricate relationships within the dataset, and their comparative performance provides valuable insights into the diverse predictive capabilities of these models in the context of CRCP spalling. [ABSTRACT FROM AUTHOR]

This study addresses the challenges of hydrogen gas detection in pipelines, focusing on the highly flammable nature and low ignition energy of hydrogen. It highlights the limitations of traditional point-based detection methods and emphasizes the need for advanced safety technologies. An integrated approach combining onsite monitoring with Internet of Things (IoT) technology is proposed to enhance systematic safety management and quick leak detection in hydrogen infrastructures. The study reviews various hydrogen detection methods, emphasizing distributed temperature sensing (DTS) techniques to identify leaks through temperature variations. Over nine months of monitoring, the results indicate that DTS temperature variations are more influenced by sensor location than chamber configuration, suggesting that external DTS chamber installation is more effective for leak detection. Additionally, the study integrates Gaussian Process Regression with Machine Learning to predict temperature distribution in pipelines, providing valuable insights for future hydrogen leak detection research. • Developed new DTS technology with enhanced resolution for sensitive hydrogen leak detection. • Confirmed DTS system's reliability against IEC 61757-2-2 standards through rigorous testing. • Integrated IoT and 3D mapping in software for efficient hydrogen leak monitoring. • Found DTS temperature variations more influenced by installation location than chamber design. • Employed GPR and Machine Learning for predicting hydrogen pipeline temperature distribution. [ABSTRACT FROM AUTHOR]

In recent years, the application of machine learning methods in the prediction of hydrological processes such as precipitation has been widely considered. These methods can analyze large volumes of data and detect the existing trends and patterns. Therefore, in the present study, machine learning methods, including random forests (RF), Kstar algorithm and Gaussian process regression (GPR), were used to predict the precipitation of Sindh River basin in India during period of 1901 to 2020. In the next step, three distinct input scenarios include (i) using monthly precipitation data and considering the memory of time series up to 5 months delay, (ii) adding periodic term to the first scenario inputs and (iii) decomposing the data using the Daubechies 4 wavelet function and creating hybrid wavelet-learning machine (W-ML) models, were prepared and introduced to the models. The performance of each method was evaluated using the root mean square error (RMSE), mean absolute error (MAE), Kling–Gupta efficiency score (KGE) and Willmott index (WI). The results showed that single models with the first scenario inputs (without taking into account the periodicity of the data) did not have good accuracy, but by adding the periodicity, the performance of these models was significantly improved, and the average value of KGE index for all studied stations increased from 0.466 to 0.672. It was also found that the GPR model for all stations could not have good performance and RF and Kstar models are the most appropriate methods for predicting precipitation in the Sindh River basin, respectively. With the application of the third scenario and the development of W-ML hybrid models, the accuracy of precipitation forecasting was significantly improved, especially the maximum precipitation values were estimated with higher accuracy than standalone models. [ABSTRACT FROM AUTHOR]

This study highlights drought forecasting for understanding the semi-arid area in India, where drought phenomena play vital role in the irrigation, drinking water supplies, and sustaining the ecological with economic balance for every nation. Therefore, drought forecasting is important for the future drought planning based on the machine learning (ML) models. Hence, The Standardized Precipitation Index (SPI) at 3- and 6-month periods have been selected and used for future drought forecasting scenarios in area. The combinations of ten inputs SPI-1- and SPI-10 were used for predicting modeling for SPI-3 and SPI-6 timescales, that modeling developed based on the historical SPI datasets from 1989 to 2019 years. The SPI-3 and SPI-6 maximum and minimum values are shown SPI-3 (2.03 and -5.522) and SPI-6 (1.94 and -6.93). The SPI is a popular method for estimating the drought analysis and has been used everywhere at global level. The developed models have been compared with each other, with the best combination of input variables selected using subset regression models and sensitivity studies. After that, the active input parameters were used for forecasting of SPI-3 and SPI-6 values to understanding of drought in semi-arid area. The finest input variables combination have been used in the Ml models and established the novel five models such as robust linear regression, bagged trees, boosted trees, support vector regression (SVM-Linear), and Matern Gaussian Process Regression (Matern GPR) models. Such kind of models first time has been applied for the forecasting of future drought conditions. Whole models were fine and improved modeling by using hyperparameters tuning, bagging, and boosting models. Entire ML models' accuracy was compared using different statistical metrics. Compared with five ML models accuracy, we have found that the Matern GPR model better accuracy than other ML models. The best model accuracy is R2 = 0.95 and 0.93, RMSE, MSE, MAE, MARE, and NSE values, respectively, for predicting SPI-3 and SPI-6 values in the area. Therefore, the Matern GPR model was identified as the finest ML algorithm for predicting SPI-3 and SPI-6 associated with other algorithms. This research demonstrates the Matern GPR model's efficacy in predicting multiscale SPI-3 and SPI-6 under climate variations. It can be helpful in soil and water resource conservation planning and management and understanding droughts in the entire basin areas of the country India. [ABSTRACT FROM AUTHOR]

AbstractOne of the most common strategies to achieve optimum designs for physical and computer high-dimensional experiments with no model pre-specification and limited budget is to uniformly spread the design points over the experimental domain, which are called uniform designs. A uniform design provides good design space coverage (space-filling), resulting in more accurate estimates using fewer points. Algorithmic search is commonly used for finding such designs, but this approach becomes ineffective for large problems. Though the construction of uniform designs theoretically is challenging, intriguing and encouraging findings have recently been accomplished for highly specialized cases. This paper gives a new theoretical construction method for high-dimensional experiments with a mixture of four-level and sixteen-level factors via an adjusted version of the multiple quadrupling algorithm (Elsawah, Communication in Mathematics and Statistics 10:623-652, 2022). The new technique (called AMQA) is simple to use since it only needs small suitable four-level designs, which may be easily generated theoretically or by algorithmic search. The quality of the small designs ensures the quality of the large designs. The effectiveness of the AMQA is investigated both theoretically and computationally. The main findings demonstrate that, compared with other widely used techniques, the AMQA provides more accurate results for high-dimensional screening using a variety of models, including polynomial, spline, and kriging models, as well as variable selection techniques, including AIC, BIC, RIC, SCAD, and LASSO. To further improve performance, a hybrid algorithm that combines the AMQA with the iterative threshold accepting algorithm is proposed. [ABSTRACT FROM AUTHOR]

Air pollution is a significant and pressing environmental and public health concern in urban areas, primarily driven by road transport. By gaining a deeper understanding of how traffic dynamics influence air pollution, policymakers and experts can design targeted interventions to tackle these critical issues. In order to analyse this relationship, a series of regression algorithms were developed utilizing the Google Project Air View (GPAV) and Dublin City's SCATS data, taking into account various spatiotemporal characteristics such as distance and weather. The analysis showed that Gaussian Process Regression (GPR) mostly outperformed Support Vector Regression (SVR) for air quality prediction, emphasizing its suitability and the importance of considering spatial variability in modelling. The model describes the data best for particulate matter (PM2.5) emissions, with R-squared (R2) values ranging from 0.40 to 0.55 at specific distances from the centre of the study area based on the GPR model. The visualization of pollutant concentrations in the study area also revealed an association with the distance between intersections. While the anticipated direct correlation between vehicular traffic and air pollution was not as pronounced, it underscores the complexity of urban emissions and the multitude of factors influencing air quality. This revelation highlights the need for a multifaceted approach to policymaking, ensuring that interventions address a broader spectrum of emission sources beyond just traffic. This study advances the current knowledge on the dynamic relationship between urban traffic and air pollution, and its findings could provide theoretical support for traffic planning and traffic control applicable to urban centres globally. [ABSTRACT FROM AUTHOR]

Molecular dynamics (MD) simulations can reduce our need for experimental tests and provide detailed insight into the chemical reactions and binding kinetics. There are two challenges while dealing with MD simulations: one is the time and length scale limitations, and the latter is efficiently processing the massive amount of data resulting from the MD simulations and generating the proper reaction rates. In this work, we evaluated the use of regression machine learning (ML) methods to solve these two challenges by developing a framework for ethanol adsorption on an Aluminium (Al) slab. This framework comprises three main stages: first, an all-atom molecular dynamics model; second, ML regression models; and third, validation and testing. In stage one, the adsorption of ethanol molecules on the Al surface for various temperatures, velocities and concentrations is simulated using the large-scale atomic/molecular massively parallel simulator (LAMMPS) and ReaxFF. The outcome of stage one is utilised for training, testing, and validating the predictive models in stages two and three. We developed and evaluated 28 different ML models for predicting the number of adsorbed molecules over time, including linear regression, support vector machine (SVM), decision trees, ensemble, Gaussian process regression (GPR), neural network (NN) and Bayesian hyper-parameter optimisation models. Based on the results, the Bayesian-based GPR showed the highest accuracy and the lowest training time. The developed model can predict the number of adsorbed molecules for new cases within seconds, while MD simulations take a few weeks. This adsorption rate can then be used in macroscale simulations to tackle the time and length scale limitations. The proposed numerical framework has the potential to be generalised and, therefore, contribute to future low-cost binding reaction estimations, providing a valuable tool for industry and experimentalists. [ABSTRACT FROM AUTHOR]

This study investigates the shear strength prediction of Fibre-Reinforced Concrete (FRC) beams reinforced with Fibre-Reinforced Polymer (FRP) bars and without stirrups through the application of Machine Learning (ML) techniques. The utilisation of FRP bars, particularly Glass Fibre-Reinforced Polymer (GFRP) and Basalt Fibre-Reinforced Polymer (BFRP) bars, has emerged as a promising alternative to traditional steel reinforcement due to their superior mechanical properties and corrosion resistance. Moreover, the incorporation of macro- and micro-discrete fibres into concrete compositions enhances both shear and flexural behaviour while reducing crack propagation induced by tensile stresses. The primary objective of this research is to develop accurate predictive models for the shear capacity of FRP-reinforced concrete beams, considering various influential parameters. Six different ML techniques, namely Multiple Linear Regression (MLR), Gaussian Process Regression (GPR), k-Nearest Neighbors (KNN), Support Vector Machines (SVMs), Artificial Neural Networks (ANNs) and Random Forest Regression (RFR), are employed to analyse the complex interactions between input variables, such as fibre type, bar diameter, concrete compressive strength, beam dimensions and the resulting shear strength. By leveraging these computational approaches, we aim to overcome the limitations of conventional analytical methods and provide robust predictions for structural design and optimisation. [ABSTRACT FROM AUTHOR]

Introduction Computational head injury models are promising tools for understanding and predicting traumatic brain injuries. However, most available head injury models are "average" models that employ a single set of head geometry (e.g. 50th-percentile U.S. male) without considering variability in these parameters across the human population. A significant variability of head shapes exists in U.S. Army soldiers, evident from the Anthropometric Survey of U.S. Army Personnel (ANSUR II). The objective of this study is to elucidate the effects of head shape on the predicted risk of traumatic brain injury from computational head injury models. Materials and Methods Magnetic resonance imaging scans of 25 human subjects are collected. These images are registered to the standard MNI152 brain atlas, and the resulting transformation matrix components (called head shape parameters) are used to quantify head shapes of the subjects. A generative machine learning model is used to generate 25 additional head shape parameter datasets to augment our database. Head injury models are developed for these head shapes, and a rapid injurious head rotation event is simulated to obtain several brain injury predictor variables (BIPVs): Peak cumulative maximum principal strain (CMPS), average CMPS, and the volume fraction of brain exceeding an injurious CMPS threshold. A Gaussian process regression model is trained between head shape parameters and BIPVs, which is then used to study the relative sensitivity of the various BIPVs on individual head shape parameters. We distinguish head shape parameters into 2 types: Scaling components |${T_{xx}}$|⁠ , |${T_{yy}}$|⁠ , and |${T_{zz}}$| that capture the breadth, length, and height of the head, respectively, and shearing components (⁠|${T_{xy}},{T_{xz}},{T_{yx}},{T_{yz}},{T_{zx}}$|⁠ , and |${T_{zy}}$|⁠) that capture the relative skewness of the head shape. Results An overall positive correlation is evident between scaling components and BIPVs. Notably, a very high, positive correlation is seen between the BIPVs and the head volume. As an example, a 57% increase in peak CMPS was noted between the smallest and the largest investigated head volume parameters. The variation in shearing components |${T_{xy}},{T_{xz}},{T_{yx}},{T_{yz}},{T_{zx}}$|⁠ , and |${T_{zy}}$| on average does not cause notable changes in the BIPVs. From the Gaussian process regression model, all 3 BIPVs showed an increasing trend with each of the 3 scaling components, but the BIPVs are found to be most sensitive to the height dimension of the head. From the Sobol sensitivity analysis, the |${T_{zz}}$| scaling parameter contributes nearly 60% to the total variance in peak and average CMPS; |${T_{yy}}$| contributes approximately 20%, whereas |${T_{xx}}$| contributes less than 5%. The remaining contribution is from the 6 shearing components. Unlike peak and average CMPS, the VF-CMPS BIPV is associated with relatively evenly distributed Sobol indices across the 3 scaling parameters. Furthermore, the contribution of shearing components on the total variance in this case is negligible. Conclusions Head shape has a considerable influence on the injury predictions of computational head injury models. Available "average" head injury models based on a 50th-percentile U.S. male are likely associated with considerable uncertainty. In general, larger head sizes correspond to greater BIPV magnitudes, which point to potentially a greater injury risk under rapid neck rotation for people with larger heads. [ABSTRACT FROM AUTHOR]

The integration of chemistry poses a major bottleneck in numerical combustion modelling, as a significant amount of simulation time is consumed in the direct integration (DI) of differential equations into thermochemistry modelling. In this work, a probabilistic machine learning (ML) framework, using Gaussian processes (GPs), is developed as a chemical source term integrator to replace DI for the prediction of a H 2 /air auto-ignition case. In this context, two algorithms, namely Gaussian Process Regression (GPR) and Gaussian Process Autoregressive Regression (GPAR), are investigated. The training dataset is generated using zero-dimensional (0D) isobaric H 2 /air auto-ignition simulations in Cantera. To address the scalability issue of the GPs, the variational inducing points method is used. This method leverages a subset of the original data for training, allowing sparse approximations. The performances of the GPR and GPAR are compared to a standard artificial neural network (ANN) model. A priori comparison with direct integration shows that both GPR ( R t e s t 2 = 0. 997) and GPAR ( R t e s t 2 = 0. 998) outperform the ANNs ( R t e s t 2 = 0. 988) by capturing latent dynamics of the chemistry when working with small datasets. Additionally, GP models offer the capability to quantify the uncertainty of each prediction, providing deeper insights into the model's limitations. It is also shown that the inference with GP-based models is slower than ANNs with speed-up factors of 1.9-2.1 relative to the 0D reactor model, whereas the ANN speed-up factor goes up to 3.0. [Display omitted] • First use of a probabilistic ML framework for chemical source term prediction. • Gaussian Processes quantify uncertainty in predicted thermo-chemical states. • Probabilistic framework's strengths and limitations compared to neural networks. • Framework surpasses neural networks in data efficiency and accuracy. [ABSTRACT FROM AUTHOR]

This paper introduces a metamodelling technique that employs gradient-enhanced Gaussian process regression (GPR) to emulate diverse internal energy densities based on the deformation gradient tensor F and electric displacement field D 0 . The approach integrates principal invariants as inputs for the surrogate internal energy density, enforcing physical constraints like material frame indifference and symmetry. This technique enables accurate interpolation of energy and its derivatives, including the first Piola-Kirchhoff stress tensor and material electric field. The method ensures stress and electric field-free conditions at the origin, which is challenging with regression-based methods like neural networks. The paper highlights that using invariants of the dual potential of internal energy density, i.e., the free energy density dependent on the material electric field E 0 , is inappropriate. The saddle point nature of the latter contrasts with the convexity of the internal energy density, creating challenges for GPR or Gradient Enhanced GPR models using invariants of F and E 0 (free energy-based GPR), compared to those involving F and D 0 (internal energy-based GPR). Numerical examples within a 3D Finite Element framework assess surrogate model accuracy across challenging scenarios, comparing displacement and stress fields with ground-truth analytical models. Cases include extreme twisting and electrically induced wrinkles, demonstrating practical applicability and robustness of the proposed approach. [ABSTRACT FROM AUTHOR]

Kriging metamodeling (also called Gaussian Process regression) is a popular approach to predict the output of a function based on few observations. The Kriging method involves length-scale hyperparameters whose optimization is essential to obtain an accurate model and is typically performed using maximum likelihood estimation (MLE). However, for high-dimensional problems, the hyperparameter optimization is problematic and often fails to provide correct values. This is especially true for Kriging-based design optimization where the dimension is often quite high. In this article, we propose a method for building high-dimensional surrogate models which avoids the hyperparameter optimization by combining Kriging sub-models with randomly chosen length-scales. Contrarily to other approaches, it does not rely on dimension reduction techniques and it provides a closed-form expression for the model. We present a recipe to determine a suitable range for the sub-models length-scales. We also compare different approaches to compute the weights in the combination. We show for a high-dimensional test problem and a real-world application that our combination is more accurate than the classical Kriging approach using MLE. [ABSTRACT FROM AUTHOR]

Highlights: IME technology is combined with conductive PET and LSR polymer materials to preform Utah arrays for BCI. The effects of three key factors on the nodal displacement were analyzed using the Kriging model. The heat transfer during injection molding was further analyzed by PDEs in order to determine the parameters more accurately. The allowable variation range of the wire diameter is obtained by the relation between the wire diameter and the current. Injection molding is an efficient and precise manufacturing technology that is widely used in the production of plastic products. In recent years, injection molding technology has made significant progress, especially with the combination of in-mold electronics (IME) technology, which makes it possible to embed electronic components directly into the surface of a product. IME technology improves the integration and performance of a product by embedding conductive materials and functional components in the mold. Brain–computer interfaces (BCIs) are a rapidly growing field of research that aims to capture, analyze, and feedback brain signals by directly connecting the brain to external devices. The Utah array, a high-density microelectrode array, has been widely used for the recording and transmission of brain signals. However, the traditional fabrication method of the Utah array suffers from high cost and low integration, which limits its promotion in practical applications. The lines that receive EEG signals are one of the key parts of a brain–computer interface system. The optimization of injection molding parameters is particularly important in order to effectively embed these lines into thin films and to ensure the precise displacement of the line nodes and the stability of signal transmission during the injection molding process. In this study, a method based on the Kriging prediction model and sparse regression partial differential equations (PDEs) is proposed to optimize the key parameters in the injection molding process. This method can effectively predict and control the displacement of nodes in the film, ensure the stability and reliability of the line during the injection process, and improve the accuracy of EEG signal transmission and system performance. The optimal injection parameters were finally obtained: a holding pressure of 525 MPa, a holding time of 50 s, and a melting temperature of 285 °C. Under this condition, the average node displacement of UA was reduced from the initial 0.19 mm to 0.89 µm, with an optimization rate of 95.32%. [ABSTRACT FROM AUTHOR]

This paper proposes an efficient Kriging-assisted improved cross-entropy (ICE) method for random-interval hybrid reliability analysis. This method employs Kriging to substitute for the projection outlines of limit state surface and incrementally updates the Kriging model within the final layer samples of ICE. A novel learning function that combines both the lower and upper confidence bounds for extremum searching is proposed to identify samples on the projection outlines. Additionally, an error-based stopping criterion (EBSC) is proposed to avoid unnecessary updates to the Kriging model. The efficiency and effectiveness of the proposed method are demonstrated through four benchmark examples. Furthermore, the method is applied to two practical engineering scenarios: the strength reliability analysis of a bogie and the resonance reliability analysis of an axially functionally graded material (FGM) pipeline. The results indicate that the proposed method achieves high levels of efficiency and accuracy. [ABSTRACT FROM AUTHOR]

This work provides novel insights into the uncertainty quantification and reliability sensitivity analysis of a direct-operated relief valve undergoing vibration. A reliability model for pressure tolerance failure of the relief valve is developed based on the criterion that the pressure tolerance exceeded the allowable value, incorporating the vibration and multiparameter uncertainty. The Simulink simulation is employed to analyze the nonlinear behaviors of the valve. To avoid excessive nonlinear calculations, a Kriging-based strategy is adopted for reliability sensitivity analysis, allowing for rapid estimation of nonlinear responses. The results are compared with those obtained by MCS to verify the performance of the proposed method. The sensitivity results indicate that the proposed method saves 98 % of computational time without much loss of accuracy. Furthermore, more attention should be paid to the spring stiffness, valve element mass, and controlled chamber volume in the structural reliability design of relief valves. [ABSTRACT FROM AUTHOR]

Background: Dementia prevalence is expected to increase as populations grow and age. Therefore, additional resources will be needed to meet the global demand for care for Alzheimer's disease and related dementias (ADRD). Objective: Estimate global and country-level health care spending attributable to ADRD and the cost of informal care for people living with ADRD. Methods: We gathered data from three systematic literature reviews and the Global Burden of Disease 2019 study. We used spatiotemporal Gaussian process regression to impute estimates for the many countries without underlying data. We projected future costs to 2050 based on past trends in costs, diagnosis rates, and institutionalization rate. Results: We estimated that in 2019, the direct health care spending attributable to ADRD across 204 countries reached $260.6 billion (95% uncertainty interval [UI] 131.6–420.4) and the cost of informal ADRD care was $354.1 billion (95% UI 190.0–544.1). On average, informal care represents 57% (95% UI 38–75%) of the total cost of care. We estimated that direct health care spending attributable to ADRD will reach $1.6 trillion (95% UI 0.6–3.3) in 2050, or 9.4% (95% UI 3.9–19.6%) of projected health spending worldwide. We estimated the cost of informal care will reach $0.9 trillion (95% UI 0.3–1.7) in 2050. Conclusions: These cost estimates underscore the magnitude of resources needed to ensure sufficient resources for people living with ADRD and highlight the role that informal care plays in provision of their care. Incorporating informal care cost estimates is critical to capture the social cost of ADRD. [ABSTRACT FROM AUTHOR]

Network real-time kinematic (NRTK) coverage is defined as the area inside a station network. In conventional NRTK, the distance between stations is limited to 100 km, thus restricting the coverage of NRTK. In this study, we propose the utilization of an ionospheric-free combination and the application of a Kriging weighting model to mitigate tropospheric delay to extend the coverage of NRTK through network expansion. A network with station distances exceeding 100 km was constructed, and the residual errors, along with the success-fix rate of integer ambiguities, were analyzed on both sunny and rainy days to confirm the potential for network expansion using the proposed method. The results confirm that the success-fix rate increased by up to 44.3% on rainy days, compared with that of the traditional interpolation method. Furthermore, a high level of performance in integer ambiguity resolution can be maintained within the expanded network, regardless of the weather conditions. [ABSTRACT FROM AUTHOR]

The turbulent ionosphere causes phase shifts to incoming radio waves on a broad range of temporal and spatial scales. When an interferometer is not sufficiently calibrated for the direction-dependent ionospheric effects, the time-varying phase shifts can cause the signal to decorrelate. The ionosphere's influence over various spatiotemporal scales introduces a baseline-dependent effect on the interferometric array. We study the impact of baseline-dependent decorrelation on high-redshift observations with the Low Frequency Array (LOFAR). Data sets with a range of ionospheric corruptions are simulated using a thin-screen ionosphere model, and calibrated using the state-of-the-art LOFAR epoch of reionization pipeline. For the first time ever, we show the ionospheric impact on various stages of the calibration process including an analysis of the transfer of gain errors from longer to shorter baselines using realistic end-to-end simulations. We find that direction-dependent calibration for source subtraction leaves excess power of up to two orders of magnitude above the thermal noise at the largest spectral scales in the cylindrically averaged autopower spectrum under normal ionospheric conditions. However, we demonstrate that this excess power can be removed through Gaussian process regression, leaving no excess power above the 10 per cent level for a |$5~$| km diffractive scale. We conclude that ionospheric errors, in the absence of interactions with other aggravating effects, do not constitute a dominant component in the excess power observed in LOFAR epoch of reionization observations of the North Celestial Pole. Future work should therefore focus on less spectrally smooth effects, such as beam modelling errors. [ABSTRACT FROM AUTHOR]

This work is focused on finding G‐optimal designs theoretically for kriging models with two‐dimensional inputs and separable exponential covariance structures. For design comparison, the notion of evenness of two‐dimensional grid designs is developed. The mathematical relationship between the design and the supremum of the mean squared prediction error (SMSPE) function is studied and then optimal designs are explored for both prospective and retrospective design scenarios. In the case of prospective designs, the new design is developed before the experiment is conducted and the regularly spaced grid is shown to be the G‐optimal design. Retrospective designs are constructed by adding or deleting points from an already existing design. Deterministic algorithms are developed to find the best possible retrospective designs (which minimizes the SMSPE). It is found that a more evenly spread design under the G‐optimality criterion leads to the best possible retrospective design. For all the cases of finding the optimal prospective designs and the best possible retrospective designs, both frequentist and Bayesian frameworks have been considered. The proposed methodology for finding retrospective designs is illustrated with a spatiotemporal river water quality monitoring experiment. [ABSTRACT FROM AUTHOR]

Machine learning as a data‐driven solution has been widely applied in the field of fatigue lifetime prediction. In this paper, three models for wideband fatigue life prediction are built based on three machine learning models, that is, support vector regression (SVR), Gaussian process regression (GPR), and artificial neural network (ANN). All the three prediction models use the parameter b of the well‐known Tovo–Benasciutti (TB) model as their outputs to realize fatigue life prediction and their generalization abilities are enhanced by employing numerous power spectrum samples with different bandwidth parameters and a variety of material properties related to fatigue life. Sufficient Monte Carlo numerical simulations demonstrate that the newly developed machine learning models are superior to the traditional frequency‐domain models in terms of life prediction accuracy and the ANN model has the best overall performance among the three developed machine learning models. Highlights: Three frequency‐domain models are devised based on machine learning.SVR, GPR, and ANN models are employed for random fatigue life predictions.Three models are suitable for predicting fatigue life under wideband random loads.Three models outperform existing frequency‐domain models in prediction accuracy. [ABSTRACT FROM AUTHOR]

Mapping of soil nutrient parameters using experimental measurements and geostatistical approaches to assist site-specific fertiliser advisories is anticipated to play a significant role in Smart Agriculture. FarmerZone is a cloud service envisioned by the Department of Biotechnology, Government of India, to provide advisories to assist smallholder farmers in India in enhancing their overall farm production. As a part of the project, we evaluated the soil spatial variability of three potato agroecological zones in India and provided soil health cards along with field-specific fertiliser recommendations for potato cultivation to farmers. Specifically, 705 surface samples were collected from three representative potato-growing districts of Indian states (Meerut, UP; Jalandhar, Punjab and Lahaul and Spiti, HP) and analysed for soil parameters such as organic carbon, macronutrients (NPK), micronutrients (Zn, Fe, Mn, and Cu), pH, and EC. The soil parameters were integrated into a geodatabase and subjected to kriging interpolation to create spatial soil maps of the targeted potato agroecological zones through best-fit experimental semivariograms. The spatial distribution showed a deficiency of soil organic carbon in two studied zones and available nitrogen among all studied zones. The available phosphorus and potassium varied among the agroecological zones. The micronutrient levels were largely sufficient in all the zones except at a few specific sites where nutrient advisories are recommended to replenish. The general management strategies were recommended based on the nutrient status in the studied area. This study clearly supports the significance of site-specific soil analytics and interpolated spatial soil mapping over any targeted agroecological zones as a promising strategy to deliver reliable advisories of fertiliser recommendations for smart farming. [ABSTRACT FROM AUTHOR]

Evaluating the reliability of deep soft rock tunnels is a very important issue to be solved. In this study, we propose a Monte Carlo simulation reliability analysis method (MCS–RAM) integrating the adaptive momentum stochastic optimization algorithm (Adam), Bayesian inference theory and Gaussian process regression (GPR) with combined kernel function, and we developed it in Python. The proposed method used the Latin hypercube sampling method to generate a dataset sample of geo-mechanical parameters, constructed combined kernel functions of GPR and used GPR to establish a surrogate model of the nonlinear mapping relationship between displacements and mechanical parameters of the surrounding rock. Adam was used to optimize the hyperparameters of the surrogate model. The Bayesian inference algorithm was used to obtain the probability distribution of geotechnical parameters and the optimal surrounding rock mechanical parameters. Finally, the failure probability was computed using MCS–RAM based on the optimized surrogate model. Through the application of an engineering case, the results indicate that the proposed method has fewer prediction errors and stronger prediction ability than Kriging or XGBoost, and it can significantly save computational time compared with the traditional polynomial response surface method. The proposed method can be used in the reliability analysis of all shapes of tunnels. [ABSTRACT FROM AUTHOR]

This study examines the challenges related to the storage and distribution of cabin materials on board during the construction of cruise ships. The construction process of cruise ships involves the assembly of various cabin materials, which are temporarily stored and transported in a complex deck environment with multiple obstacles. These processes can be viewed as location-routing problems (LRPs). However, existing solutions for LRPs do not account for the obstacles commonly encountered in deck environments. Therefore, this paper introduces a new variant of LRPs, termed as the BE-LRP (blocking environment location-routing problem). Initially, by considering the optimization objectives of minimizing the distribution distance and the number of vehicles required, the paper develops an optimization model for the location-routing of on-board materials while taking obstacles into consideration. Subsequently, a method for estimating blocked distances is proposed. This method utilizes Gaussian process regression to predict blocked distances and enables the rapid estimation of distances obstructed by obstacles between different destinations. Lastly, a hybrid obstacle blocking distance processing HO-NSGAII algorithm is formulated to address the BE-LRP. Experimental comparisons demonstrate that the distribution scheme derived from this research method does not necessitate modifications to the delivery distance when compared to the traditional NSGAII algorithm, and better aligns with the actual deck conditions. With an equivalent number of delivery vehicles, the scheme results in shorter delivery distances. The delivery distance can be reduced by 3–17% under varying circumstances. These findings are advantageous for enhancing the efficiency of cruise ship construction and mitigating the impact of warehousing and distribution on construction progress. [ABSTRACT FROM AUTHOR]

In near coastal environments, nautical charts provide crucial information for navigation and routing both in real-time operations and during planning stages. The cost of data collection as well as capacity constraints in the processing pipeline make reliable bathymetric information in such areas sparse. Prioritization rules can help guide the efforts to where information is the most valuable. AIS data provide accounts of real ship movements, indicating both desirable paths and minimum depths. We propose a statistical model for combining sparse bathymetric soundings with AIS observations for improved prediction of depths for generation of feasible transportation corridors. The method relies on viewing AIS draughts as censored observations of the true depth. A case-study is performed for the southern archipelago of Gothenburg using the program R-INLA. The non-stationarity caused by having boundaries with known (zero) depth and holes (land) in the domain is handled through discretization. Varying amounts of AIS data, ranging from none to 1824 observations, are used in the experiments. Results show predicted depths within the range of data values, and that inclusion of AIS data serve to push the field down to ensure that traverseable areas are predicted as such revealing corridors in narrow passages where bathymetric soundings are lacking. [ABSTRACT FROM AUTHOR]

Numerous hydrological applications, such as soil erosion estimation, water resource management, and rain driven damage assessment, demand accurate and reliable rainfall erosivity data. However, the scarcity of gauge rainfall records and the inherent uncertainty in satellite and reanalysis-based rainfall datasets limit rainfall erosivity assessment globally. Here, we present a new global rainfall erosivity dataset (0.1° × 0.1° spatial resolution) integrating satellite (CMORPH and IMERG) and reanalysis (ERA5-Land) derived rainfall erosivity estimates with gauge rainfall erosivity observations collected from approximately 6,200 locations across the globe. We used a machine learning-based Gaussian Process Regression (GPR) model to assimilate multi-source rainfall erosivity estimates alongside geoclimatic covariates to prepare a unified high-resolution mean annual rainfall erosivity product. It has been shown that the proposed rainfall erosivity product performs well during cross-validation with gauge records and inter-comparison with the existing global rainfall erosivity datasets. Furthermore, this dataset offers a new global rainfall erosivity perspective, addressing the limitations of existing datasets and facilitating large-scale hydrological modelling and soil erosion assessments. [ABSTRACT FROM AUTHOR]

AbstractData transformations are essential for broad applicability of parametric regression models. However, for Bayesian analysis, joint inference of the transformation and model parameters typically involves restrictive parametric transformations or nonparametric representations that are computationally inefficient and cumbersome for implementation and theoretical analysis, which limits their usability in practice. This paper introduces a simple, general, and efficient strategy for joint posterior inference of an unknown transformation and all regression model parameters. The proposed approach directly targets the posterior distribution of the transformation by linking it with the marginal distributions of the independent and dependent variables, and then deploys a Bayesian nonparametric model via the Bayesian bootstrap. Crucially, this approach delivers (1) joint posterior consistency under general conditions, including multiple model misspecifications, and (2) efficient Monte Carlo (not Markov chain Monte Carlo) inference for the transformation and all parameters for important special cases. These tools apply across a variety of data domains, including real-valued, positive, and compactly-supported data. Simulation studies and an empirical application demonstrate the effectiveness and efficiency of this strategy for semiparametric Bayesian analysis with linear models, quantile regression, and Gaussian processes. The R package SeBR is available on CRAN. [ABSTRACT FROM AUTHOR]

The present study involves the development of a new method, a range-dependent multivariate adaptive regression splines–genetic algorithm (RD-MARS-GA), for monthly streamflow prediction. The results of the RD-MARS-GA model are compared with standard MARS, Gaussian process regression, principal component regression, M5 model tree and random forest. The input data utilized in the study combine satellite and ground-based datasets. Monthly data from the Treene, Schaale, and Wandse rivers in Northern Germany were used as case studies. The satellite datasets used were sourced from Modern-Era Retrospective Analysis for Research and Applications, version 2 (MERRA-2), Integrated Multi-Satellite Retrievals for Global Precipitation Measurement (IMERG GPM) and Moderate Resolution Imaging Spectroradiometer Evapotranspiration (MODIS). A comparison of the results reveals that the RD-MARS-GA can improve the forecasting accuracy of the MARS model across all watersheds at different temporal scales. The RD-MARS-GA model surpassed the other regression models and improved the root mean square accuracy of the MARS model by 12.8%, 26.7% and 18.5% in predicting one-month-ahead streamflow of the rivers Treene, Schaale, and Wandse, respectively. [ABSTRACT FROM AUTHOR]

The state parameter (ѱ) accounts for both relative density and effective stress, which influence the cyclic stress or liquefaction characteristic of the soil significantly. This study presents a ѱ‐based probabilistic liquefaction evaluation method using six soft computing (SC) techniques. The liquefaction probability of failure (PL) is calculated using the first‐order second moment (FOSM) method based on the cone penetration test (CPT) database. Then, six SC techniques, such as Gaussian process regression (GPR), relevance vector machine (RVM), functional network (FN), genetic programming (GP), minimax probability machine regression (MPMR) and multivariate adaptive regression splines (MARS), are used to predict PL. The performance of these models is examined using nine statistical indices. Additionally, plots such as regression plots, Taylor diagrams, error matrix and rank analysis are shown to assess the SC model's performance. Finally, sensitivity analysis is performed using the cosine amplitude method (CAM) to assess the influence of input parameters on output. The current study demonstrates that SC models based on state parameter predict PL effectively. RVM and MPMR models closely follow the GPR model in terms of performance, which is superior to the other models. Notably, two equations are generated using GP and MARS models to predict PL. The results of the sensitivity analysis reveal the magnitude of earthquake (Mw) as the most sensitive parameter. The outcomes of this research will offer risk evaluations for geotechnical engineering designs and expand the use of state parameter‐based SC models in liquefaction analysis. [ABSTRACT FROM AUTHOR]

This paper aims to apply data-driven maximum power point tracking (MPPT) techniques specifically tailored for fuel cell vehicle (FCV) supported hybrid DC microgrids to enhance the power harvesting capability of fuel cell (FC) stacks. Compared to existing MPPT techniques, the current study focuses on developing and evaluating data-driven approaches for maximum power extraction by dynamically determining the operating point of FC power sources through a Zeta converter. An in-depth analysis is conducted by considering parameters such as efficiency, tracking accuracy, response time, and robustness to variations in load demand and operating conditions. The performance results validate that the developed three-layer neural network (TNN)-based MPPT gives better performance findings than Gaussian process regression (GPR), support vector regression (SVR), decision tree regression (DTR), and bagging ensemble decision tree (BEDT). In the performance evaluation phase, a vehicular FC with a rating of 1.26 kW is designed and operated within the temperature range of 320 K to 343 K for hydrogen pressure values ranging from 1 bar to 1.8 bar. For these operational conditions, the prediction accuracy value of the proposed TNN method is 99.6% while the performance values GPR, SVR, DTR, and BEDT are 99%, 98.6%, 97.2%, and 96%. In addition, system efficiency is increased by 0.98%, 1.25%, 2.51%, and 3.02% compared to GPR, SVR, DTR, and BEDT, respectively. [Display omitted] • Developing data-driven algorithms-based MPPT techniques for vehicular FCs. • Providing advantages regarding high-speed tracking capability and tracking accuracy. • Providing stable DC bus voltage operation. • Improving system performance under dynamic load changes and operating conditions. [ABSTRACT FROM AUTHOR]

The incorporation of carbon nanotubes (CNTs) in concrete can improve the physical, mechanical, and durability properties. However, the interaction of CNTs with concrete and their effect on the mechanical properties remains a challenging issue. Also, the determination of mechanical properties through experimental testing is time‐consuming, laborious, and uneconomical. This study focuses on the development of machine learning (ML) models for the prediction of the mechanical properties of concrete. A comprehensive data set of 758 CNT‐modified concrete specimens was established for the compressive strength (CS), split tensile strength (STS), flexural strength (FS), and modulus of elasticity (MOE) values from the experimental studies in the literature. Afterward, the predictive models were developed using multilinear regression (MLR), support vector machine (SVM), ensemble methods (EN), regression tree (RT), and Gaussian process regression (GPR). It was found that among ML models, the GPR model predicted the CS, STS, and FS at the highest efficiency with the coefficient of determination (R2) of 0.83, 0.78, and 0.93, respectively while the performance of the SVM model was superior for predicting MOE with an R2 value of 0.91. The mean absolute error (MAE) of the GPR model for CS, STS, FS, and MOE were 2.92, 0.26, 0.35, and 1.31, respectively which were also lesser than other models. The training time of different models demonstrated that the GPR model has also a lower training time (~3 s) as compared to other models which indicates it has a high accuracy‐to‐time cost ratio. Further, the most influential parameters on CS were age, cement, water–cement ratio, and carbon nanotubes. The one‐way partial dependence analysis showed a direct correlation for age and cement but an inverse correlation for the water–cement ratio and fine aggregate. The graphical user interface provides the implication of the developed models for practical applications. [ABSTRACT FROM AUTHOR]

AbstractOccupational exposure to particulate matter (PM) can result in multiple adverse health effects and should be minimized to protect workers’ health. PM exposure at the workplace can be complex with many potential sources and fluctuations over time, making it difficult to control. Dynamic maps that visualize how PM is distributed throughout a workplace over time can help in gaining better insights into when and where exposure occurs. This study explored the use of spatiotemporal modeling followed by kriging for the development of dynamic PM concentration maps in an experimental setting and a workplace setting. Data was collected using personal low-cost PM sensors and an indoor location tracking system, mounted on a moving robot or worker. Maps were generated for an experimental study with one simulated robot worker and a workplace study with four workers. Cross-validation was performed to evaluate the performance and robustness of three types of spatiotemporal models (metric, separable, and summetric) and, as an additional external validation, model estimates were compared with measurements from sensors that were placed stationary in the laboratory or workplace. Spatiotemporal models and maps were generated for both the experimental and workplace studies, with average root mean squared error (RMSE) from 10-fold cross-validation ranging from 7–12 and 73–127 µg/m3, respectively. Workplace models were relatively more robust compared to the experimental study (relative SD ranging from 8–14% of the average RMSE vs. 27–56%, respectively), presumably due to the larger number of parallel measurements. Model estimates showed low to moderate fits compared to stationary sensor measurements (R2 ranging from 0.1–0.5), indicating maps should be interpreted with caution and only used indicatively. Together, these findings show the feasibility of using spatiotemporal modeling for generating dynamic concentration maps based on personal data. The described method could be applied for exposure characterization within comparable study designs or can be expanded further, for example by developing real-time, location-based worker feedback systems, as efficient tools to visualize and communicate exposure risks. [ABSTRACT FROM AUTHOR]

This paper proposes a new stochastic model updating method based on the homotopy surrogate model (HSM) and Bayesian sampling. As a novel intrusive surrogate model, the HSM is established by the homotopy stochastic finite element (FE) method. Then combining the advanced delayed‐rejection adaptive Metropolis–Hastings sampling technology with HSM, the structural FE model can be updated by uncertain measurement modal data. The numerical results show that the updating effectiveness of the proposed method is better than that of the Bayesian methods with the non‐intrusive surrogate models, such as stochastic response surface model and Kriging model. Compared to the Bayesian method with the intrusive second‐order perturbation model, the updating results of the proposed method are more accurate, especially when the fluctuation of the uncertain measured data is large and the stiffness of the structure significantly changes. The model updating results of a cable‐stayed bridge show that the statistic modal properties of the updated bridge model have a very good agreement with the uncertain measurement modal data. [ABSTRACT FROM AUTHOR]

With the increasing use of data-driven modeling methods, new approaches to complex problems in the field of wind energy can be addressed. Topics reviewed through the literature include wake modeling, performance monitoring and controls applications, condition monitoring and fault detection, and other data-driven research. The literature shows the advantages of data-driven methods: a reduction in computational expense or complexity, particularly in the cases of wake modeling and controls, as well as various data-driven methodologies' aptitudes for predictive modeling and classification, as in the cases of fault detection and diagnosis. Significant work exists for fault detection, while less work is found for controls applications. A methodology for creating data-driven wind turbine models for arbitrary performance parameters is proposed. Results are presented utilizing the methodology to create wind turbine models relating active adaptive twist to steady-state rotor thrust as a performance parameter of interest. Resulting models are evaluated by comparing root-mean-square-error (RMSE) on both the training and validation datasets, with Gaussian process regression (GPR), deemed an accurate model for this application. The resulting model undergoes particle swarm optimization to determine the optimal aerostructure twist shape at a given wind speed with respect to the modeled performance parameter, aerodynamic thrust load. The optimization process shows an improvement of 3.15% in thrust loading for the 10 MW reference turbine, and 2.66% for the 15 MW reference turbine. [ABSTRACT FROM AUTHOR]

The operational parameters of plant protection unmanned aerial vehicles (UAVs) significantly impact spraying effectiveness, but the underlying mechanism remains unclear. This paper conducted a full factorial experiment with varying flight speeds, heights, and nozzle flow rates to collect parameter space data. Using the Kriging surrogate model, we characterized this parameter space and subsequently optimized the average deposition rate and coefficient of variation by employing a variable crossover (mutation) probability multi-objective genetic algorithm. In the obtained Pareto front, the average sedimentation rate is no less than 46%, with a maximum of 56.08%, and the CV coefficient is no more than 13.91%, with a minimum of only 8.42%. These optimized parameters enhance both the average deposition rate and spraying uniformity compared to experimental data. By employing these optimized parameters in practical applications, a balance between the maximum average deposition rate and minimum coefficient of variation can be achieved during UAV spraying, thereby reducing pesticide usage, promoting sustainable agriculture, and mitigating instances of missed spraying and re-spraying. [ABSTRACT FROM AUTHOR]

Over the years, prediction techniques for the highly variable angular velocity of the Earth represented by Earth's rotation (UT1-UTC) and length-of-day (LOD) have been continuously improved. This is because many applications like navigation, astronomy, space exploration, climate studies, timekeeping, disaster monitoring, and geodynamic studies, all rely on predictions of these Earth rotation parameters. They provide early warning of changes in the Earth's rotation, allowing various industries and scientific fields to operate more precisely and efficiently. Thus, in our study, we focused on short-term prediction for UT1-UTC (dUT1) and LOD. Our prediction approach is to combine machine learning (ML) technique with efficient evolutionary computation (EC) algorithms to achieve reliable and improved predictions. Gaussian process regression (GPR) is used as the ML technique with genetic algorithm (GA) as the EC algorithm. GA is used for hyperparameter optimization of GPR model as selecting appropriate values for hyperparameter are essential to ensure that the prediction model can accurately capture the underlying patterns in the data. We conducted some experiments with our prediction approach to thoroughly test its capabilities. Moreover, two forecasting strategies were used to assess the performance in both hindcast and operational settings. In most of the experiments, the data used are the multi-technique combinations (C04) generated by International Earth Rotation and Reference Systems Service (IERS). In one of the experiments, we also investigated the performance of our prediction model on dUT1 and LOD from four different products obtained from IERS EOP 20 C04, DTRF20, JTRF20 and USNO. The prediction products are evaluated with real estimates of the EOP product with which the model is trained. The combined excitations of the atmosphere, oceans, hydrology, and sea level (AAM + OAM + HAM + SLAM) are used as predictors because they are highly correlated to the input data. The results depict the highest performance of 0.412 ms in dUT1 and 0.092 ms/day in LOD, on day 10 of predictions. It is worth noting that the later predictions were obtained by incorporating the uncertainty of the input data as weights in the prediction model, which was a novel approach tested in this study. [ABSTRACT FROM AUTHOR]

Spatial models are essential in the prediction of climate phenomena because they can model the complex relationships between different locations. In this study, we discuss an expanded spatial Durbin model with ordinary kriging on unobserved locations (ESDMOK) to predict rainfall patterns in Java Island. The classical spatial Durbin model needed to be expanded to obtain a parameter estimation for each location. We combined this with ordinary kriging because the data were not available in some locations. The data were taken from the National Aeronautics and Space Administration Prediction of Worldwide Energy Resources (NASA POWER) website. Since climate data are big data, we implement a big data analytics approach, namely the data analytics life cycle method. As the exogenous variables, we used air temperature, humidity, solar irradiation, wind speed, and surface pressure. The authors developed an R-Shiny web applications to implement our proposed technique. Using our proposed technique, we obtained more accurate and reliable climate data prediction, indicated by the mean absolute percentage error (MAPE), which was equal to 1.956%. The greatest effect on rainfall was given by the surface pressure variable, and the smallest was wind speed. [ABSTRACT FROM AUTHOR]

Chlorophyll monitoring is an important topic in phenotypic research. For fruit trees, chlorophyll content can reflect the real-time photosynthetic capacity, which is a great reference for nutrient status assessment. Traditional in situ estimation methods are labor- and time-consuming. Remote sensing spectral imagery has been widely applied in agricultural research. This study aims to explore a transferable model to estimate canopy SPAD across growth stages and tree species. Unmanned aerial vehicle (UAV) system was applied for multispectral images acquisition. The results showed that the univariate model yielded with Green Normalized Difference Vegetation Index (GNDVI) gave valuable prediction results, providing a simple and effective method for chlorophyll monitoring for single species. Reflection features (RF) and texture features (TF) were extracted for multivariate modeling. Gaussian Process Regression (GPR) models yielded better performance for mixed species research than other algorithm models, and the R² of the RF+TF+GPR model was approximately 0.7 in both single and mixed species. In addition, this method can also be used to predict canopy SPAD over various growth stages, especially in the third and fourth stages with R² higher than 0.6. This paper highlights the importance of using RF+TF for canopy feature expression and deep connection exploration between canopy features with GPR algorithm. This research provides a universal model for canopy SPAD inversion which can promote the growth status monitoring and management of fruit trees. [ABSTRACT FROM AUTHOR]