Micro-Scale Particle Tracking: From Conventional to Data-Driven Methods.

Micro-scale positioning techniques have become essential in numerous engineering systems. In the field of fluid mechanics, particle tracking velocimetry (PTV) stands out as a key method for tracking individual particles and reconstructing flow fields. Here, we present an overview of the micro-scale particle tracking methodologies that are predominantly employed for particle detection and flow field reconstruction. It covers various methods, including conventional and data-driven techniques. The advanced techniques, which combine developments in microscopy, photography, image processing, computer vision, and artificial intelligence, are making significant strides and will greatly benefit a wide range of scientific and engineering fields.

WyNDA: A method to discover mathematical models of dynamical systems from data.

This paper introduces a novel method called Wide-Array of Nonlinear Dynamics Approximation (WyNDA) for extracting mathematical models of dynamical systems from data. A key advantage of this method over existing approaches lies in its suitability for online implementation. Moreover, WyNDA stands out by not relying on optimization or machine learning, ensuring computational efficiency. The fundamental concept revolves around approximating the unknown function of a dynamical system through a diverse set of basis functions that encapsulate the available data. An adaptive observer is then employed to iteratively refine this approximation and estimate the associated parameters. The efficacy of the proposed method is demonstrated through numerical simulations encompassing linear systems, nonlinear systems, and control systems. The results underscore the method's ability to successfully unveil the governing equations of dynamical systems, highlighting its potential for extracting intricate system dynamics from observational data.•WyNDA represents a novel approach for uncovering mathematical models of dynamical systems from data.•Utilizing a series of basis functions, WyNDA effectively approximates the unknown structure inherent in dynamical systems.•The validation of WyNDA involves benchmark equations of dynamical systems, confirming its efficacy in diverse scenarios.

Prediction of Protein Aggregation Propensity via Data-Driven Approaches.

Protein aggregation occurs when misfolded or unfolded proteins physically bind together and can promote the development of various amyloid diseases. This study aimed to construct surrogate models for predicting protein aggregation via data-driven methods using two types of databases. First, an aggregation propensity score database was constructed by calculating the scores for protein structures in the Protein Data Bank using Aggrescan3D 2.0. Moreover, feature- and graph-based models for predicting protein aggregation have been developed by using this database. The graph-based model outperformed the feature-based model, resulting in an R2 of 0.95, although it intrinsically required protein structures. Second, for the experimental data, a feature-based model was built using the Curated Protein Aggregation Database 2.0 to predict the aggregated intensity curves. In summary, this study suggests approaches that are more effective in predicting protein aggregation, depending on the type of descriptor and the database.

A data-driven approach to categorize patients with traumatic spinal cord injury: cluster analysis of a multicentre database.

Background: Conducting clinical trials for traumatic spinal cord injury (tSCI) presents challenges due to patient heterogeneity. Identifying clinically similar subgroups using patient demographics and baseline injury characteristics could lead to better patient-centered care and integrated care delivery. Purpose: We sought to (1) apply an unsupervised machine learning approach of cluster analysis to identify subgroups of tSCI patients using patient demographics and injury characteristics at baseline, (2) to find clinical similarity within subgroups using etiological variables and outcome variables, and (3) to create multi-dimensional labels for categorizing patients. Study design: Retrospective analysis using prospectively collected data from a large national multicenter SCI registry. Methods: A method of spectral clustering was used to identify patient subgroups based on the following baseline variables collected since admission until rehabilitation: location of the injury, severity of the injury, Functional Independence Measure (FIM) motor, and demographic data (age, and body mass index). The FIM motor score, the FIM motor score change, and the total length of stay were assessed on the subgroups as outcome variables at discharge to establish the clinical similarity of the patients within derived subgroups. Furthermore, we discussed the relevance of the identified subgroups based on the etiological variables (energy and mechanism of injury) and compared them with the literature. Our study also employed a qualitative approach to systematically describe the identified subgroups, crafting multi-dimensional labels to highlight distinguishing factors and patient-focused insights. Results: Data on 334 tSCI patients from the Rick Hansen Spinal Cord Injury Registry was analyzed. Five significantly different subgroups were identified (p-value ≤0.05) based on baseline variables. Outcome variables at discharge superimposed on these subgroups had statistically different values between them (p-value ≤0.05) and supported the notion of clinical similarity of patients within each subgroup. Conclusion: Utilizing cluster analysis, we identified five clinically similar subgroups of tSCI patients at baseline, yielding statistically significant inter-group differences in clinical outcomes. These subgroups offer a novel, data-driven categorization of tSCI patients which aligns with their demographics and injury characteristics. As it also correlates with traditional tSCI classifications, this categorization could lead to improved personalized patient-centered care.

A Hybrid Model for Vehicle Acceleration Prediction.

Accurate prediction of vehicle acceleration has significant practical applications. Deep learning, as one of the methods for acceleration prediction, has shown promising applications in acceleration prediction. However, due to the influence of multiple factors on acceleration, a single data model may not be suitable for various driving scenarios. Therefore, this paper proposes a hybrid approach for vehicle acceleration prediction by combining clustering and deep learning techniques. Based on historical data of vehicle speed, acceleration, and distance to the preceding vehicle, the proposed method first clusters the acceleration patterns of vehicles. Subsequently, different prediction models and parameters are applied to each cluster, aiming to improve the prediction accuracy. By considering the unique characteristics of each cluster, the proposed method can effectively capture the diverse acceleration patterns. Experimental results demonstrate the superiority of the proposed approach in terms of prediction accuracy compared to benchmarks. This paper contributes to the advancement of sensor data processing and artificial intelligence techniques in the field of vehicle acceleration prediction. The proposed hybrid method has the potential to enhance the accuracy and reliability of acceleration prediction, enabling applications in various domains, such as autonomous driving, traffic management, and vehicle control.

Quantitative Representation of Water Quality Biotoxicity by Algal Photosynthetic Inhibition.

The method based on the photosynthetic inhibition effect of algae offers the advantages of swift response and straightforward measurement. Nonetheless, this effect is influenced by both the environment and the state of the algae themselves. Additionally, a single parameter is vulnerable to uncertainties, rendering the measurement accuracy and stability inadequate. This paper employed currently utilized photosynthetic fluorescence parameters, including Fv/Fm(maximum photochemical quantum yield), Performance Indicator (PIabs), Comprehensive Parameter Index (CPI) and Performance Index of Comprehensive Toxicity Effect (PIcte), as quantitative toxicity characteristic parameters. The paper compared the univariate curve fitting results with the multivariate data-driven model results and investigated the effectiveness of Back Propagation(BP) Neural Network and Support Vector Machine for Regression (SVR) models to enhance the accuracy and stability of toxicity detection. Using Dichlorophenyl Dimethylurea (DCMU) samples as an example, the mean Relative Root Mean Square Error (RRMSE) corresponding to the optimal parameter PIcte for the dose-effect curve fitting was 1.246 in the concentration range of 1.25-200 µg/L. On the other hand, the mean RRMSEs corresponding to the results of the BP neural network and SVR models were 0.506 and 0.474, respectively. Notably, BP neural network exhibited excellent prediction accuracy in the medium-high concentration range of 7.5-200 µg/L, with a mean RRSME of only 0.056. Regarding the stability of the results, the mean Relative Standard Deviation (RSD) of the univariate dose-effect curve results was 15.1% within the concentration range of 50-200 µg/L. In contrast, the mean RSDs for both BP neural network and SVR results were less than 5%. In the concentration range of 1.25-200 µg/L, the mean RSDs were 6.1% and 16.5%, with the BP neural network performing well. The experimental results of Atrazine were analyzed to further validate the effectiveness of the BP neural network in improving the accuracy and stability of results. These findings provided valuable insights for the development of biotoxicity detection by using the algae photosynthetic inhibition method.

MMINP: A computational framework of microbe-metabolite interactions-based metabolic profiles predictor based on the O2-PLS algorithm.

The gut metabolome acts as an intermediary between the gut microbiota and host, and has tremendous diagnostic and therapeutic potential. Several studies have utilized bioinformatic tools to predict metabolites based on the different aspects of the gut microbiome. Although these tools have contributed to a better understanding of the relationship between the gut microbiota and various diseases, most of them have focused on the impact of microbial genes on the metabolites and the relationship between microbial genes. In contrast, relatively little is known regarding the effect of metabolites on the microbial genes or the relationship between these metabolites. In this study, we constructed a computational framework of Microbe-Metabolite INteractions-based metabolic profiles Predictor (MMINP), based on the Two-Way Orthogonal Partial Least Squares (O2-PLS) algorithm to predict the metabolic profiles associated with gut microbiota. We demonstrated the predictive value of MMINP relative to that of similar methods. Additionally, we identified the features that would profoundly impact the prediction performance of data-driven methods (O2-PLS, MMINP, MelonnPan, and ENVIM), including the training sample size, host disease state, and the upstream data processing methods of the different technical platforms. We suggest that when using data-driven methods, similar host disease states and preprocessing methods, and a sufficient number of training samples are necessary to achieve accurate prediction.

MMINP fully considers internal and mutual correlations in metabolites and microbial genes and infers metabolite information through their real joint parts.The feasibility of predicting metabolic profiles using gut microbiome data should be based on the premise of similar host disease states, similar preprocessing methods, and a sufficient number of training samples.Although the accuracy of predicted specific metabolites is affected by multiple factors, the systematic conclusions presented for predicted metabolites at higher levels (e.g., class level) are accurate, allowing metabolite prediction to be applied to the discovery of potential metabolite markers.

A new data driven method for summarising multiple cause of death data.

BACKGROUND: National mortality statistics are based on a single underlying cause of death. This practice does not adequately represent the impact of the range of conditions experienced in an ageing population in which multimorbidity is common. METHODS: We propose a new method for weighting the percentages of deaths attributed to different causes that takes account of the patterns of associations among underlying and contributing causes of death. It is driven by the data and unlike previously proposed methods does not rely on arbitrary choices of weights which can over-emphasise the contribution of some causes of death. The method is illustrated using Australian mortality data for people aged 60 years or more. RESULTS: Compared to the usual method based only on the underlying cause of death the new method attributes higher percentages of deaths to conditions like diabetes and dementia that are frequently mentioned as contributing causes of death, rather than underlying causes, and lower percentages to conditions to which they are closely related such as ischaemic heart disease and cerebrovascular disease. For some causes, notably cancers, which are usually recorded as underlying causes with few if any contributing causes the new method produces similar percentages to the usual method. These different patterns among groups of related conditions are not apparent if arbitrary weights are used. CONCLUSION: The new method could be used by national statistical agencies to produce additional mortality tables to complement the current tables based only on underlying causes of death.

The Chinese Communist Party and regulatory transparency in China's food industry.

While it is widely accepted that the Chinese Communist Party (CCP) occupies a dominant position in the Chinese political system, few studies have demonstrated CCP's dominant position based on rigorous statistical analysis. Our paper presents the first such analysis using an innovative measure of regulatory transparency in the food industry across nearly 300 prefectures in China over 10 years. We show that actions by the CCP, while broadly scoped and not targeting the food industry, significantly improved regulatory transparency in the industry. In sharp contrast, food-industry-specific interventions by the State Council, which exercises direct regulatory supervision of the industry, had no impact on regulatory transparency. These results hold in various specifications and robustness checks. Our research contributes to research in China's political system by empirically and explicitly demonstrating the dominating power of the CCP.

AI-based modeling and data-driven identification of moving load on continuous beams.

Traffic load identification for bridges is of great significance for overloaded vehicle control as well as the structural management and maintenance in bridge engineering. Unlike the conventional load identification methods that always encounter problems of ill-condition and difficulties in identifying multi parameters simultaneously when solving the motion equations inversely, a novel strategy is proposed based on smart sensing combing intelligent algorithm for real-time traffic load monitoring. An array of lead zirconium titanate sensors is applied to capture the dynamic responses of a beam bridge, while the Long Short-Term Memory (LSTM) neural network is employed to establish the mapping relations between the dynamic responses of the bridge and the traffic load through data mining. The results reveal that, with the real-time strain responses fed into the LSTM network, the speed and magnitude of the moving load may be identified simultaneously with high accuracy when compared to the practically applied load. The current method may facilitate highly efficient identification of the time-varying characteristics of moving loads and may provide a useful tool for long-term traffic load monitoring and traffic control for in-service bridges.

Generating Edge Cases for Testing Autonomous Vehicles Using Real-World Data.

In the past decade, automotive companies have invested significantly in autonomous vehicles (AV), but achieving widespread deployment remains a challenge in part due to the complexities of safety evaluation. Traditional distance-based testing has been shown to be expensive and time-consuming. To address this, experts have proposed scenario-based testing (SBT), which simulates detailed real-world driving scenarios to assess vehicle responses efficiently. This paper introduces a method that builds a parametric representation of a driving scenario using collected driving data. By adopting a data-driven approach, we are then able to generate realistic, concrete scenarios that correspond to high-risk situations. A reinforcement learning technique is used to identify the combination of parameter values that result in the failure of a system under test (SUT). The proposed method generates novel, simulated high-risk scenarios, thereby offering a meaningful and focused assessment of AV systems.

Seismic Waveform Inversion Capability on Resource-Constrained Edge Devices.

Seismic full wave inversion (FWI) is a widely used non-linear seismic imaging method used to reconstruct subsurface velocity images, however it is time consuming, has high computational cost and depend heavily on human interaction. Recently, deep learning has accelerated it's use in several data-driven techniques, however most deep learning techniques suffer from overfitting and stability issues. In this work, we propose an edge computing-based data-driven inversion technique based on supervised deep convolutional neural network to accurately reconstruct the subsurface velocities. Deep learning based data-driven technique depends mostly on bulk data training. In this work, we train our deep convolutional neural network (DCN) (UNet and InversionNet) on the raw seismic data and their corresponding velocity models during the training phase to learn the non-linear mapping between the seismic data and velocity models. The trained network is then used to estimate the velocity models from new input seismic data during the prediction phase. The prediction phase is performed on a resource-constrained edge device such as Raspberry Pi. Raspberry Pi provides real-time and on-device computational power to execute the inference process. In addition, we demonstrate robustness of our models to perform inversion in the presence on noise by performing both noise-aware and no-noise training and feeding the resulting trained models with noise at different signal-to-noise (SNR) ratio values. We make great efforts to achieve very feasible inference times on the Raspberry Pi for both models. Specifically, the inference times per prediction for UNet and InversionNet models on Raspberry Pi were 22 and 4 s respectively whilst inference times for both models on the GPU were 2 and 18 s which are very comparable. Finally, we have designed a user-friendly interactive graphical user interface (GUI) to automate the model execution and inversion process on the Raspberry Pi.

Impaired feedforward control of movements in pianists with focal dystonia.

Learning accurate and fast movements typically accompanies the modulation of feedforward control. Nevertheless, it remains unclear how motor skill learning modulates feedforward control, such as through maladaptation of the sensorimotor system by extensive training (e.g., task-specific dystonia). Here, we examined the modulation of feedforward control through motor skill learning while focusing on the motion of piano playing at either a natural tempo or the fastest tempo. The current study compared the kinematics and keypress data among individuals in three groups: healthy and well-trained pianists (i.e., subjects with skill learning), non-musicians (i.e., subjects without skill learning), and patients with focal-hand dystonia (FHD) (i.e., subjects with maladaptation by skill learning). Compared to healthy pianists, patients with FHD showed impairment in some feedforward motion components that are relevant to classifying the two playing tempi. However, while focusing on motion components that are irrelevant to the tempo classification, patients with FHD showed movements comparable to those of healthy pianists. Furthermore, patients with FHD demonstrated significantly slower movement times than healthy pianists. Our results suggest that maladaptation by skill learning affects parts of feedforward control rather than its entirety. Nevertheless, the affected feedforward components are relevant to performing movements as fast as possible, which may underlie the speed dependence of dystonic symptoms.

Integrated data-driven strategy to optimize the processes configuration for full-scale wastewater treatment plant predesign.

Wastewater treatment plants (WWTPs) play an irreplaceable role in eliminating pollutants from domestic and industrial wastewater and contribute to water recycling. Nowadays, the selection of processes configuration of WWTPs mainly depends on the local wastewater treatment standards and the experience of wastewater engineers rather than an intelligent data-driven strategy. In this study, an integrated data-driven strategy consisting of t-distributed stochastic neighbor embedding (t-SNE) and deep neural networks (DNNs) is proposed for optimizing the processes configuration of full-scale WWTP predesign. A large dataset with 14,647 samples collected from 10 full-scale WWTPs with distinct treatment processes is clustered by the t-SNE method based on the influent characteristics, and four meaningful clusters (Clusters I-IV) are identified for the subsequent development of DNN classification models. All four DNN models achieve acceptable classification accuracy (>0.8975) and the maximal testing accuracy is 0.9505. The DNN models are capable of finding the optimized processes configuration of WWTPs under target scenarios. Our results highlight the strength of combining the t-SNE and the DNN models to utilize the relationships between key parameters and processes configuration of WWTPs, and help engineers predesign WWTPs with the optimal processes configuration.

Health Monitoring of Large-Scale Civil Structures: An Approach Based on Data Partitioning and Classical Multidimensional Scaling.

A major challenge in structural health monitoring (SHM) is the efficient handling of big data, namely of high-dimensional datasets, when damage detection under environmental variability is being assessed. To address this issue, a novel data-driven approach to early damage detection is proposed here. The approach is based on an efficient partitioning of the dataset, gathering the sensor recordings, and on classical multidimensional scaling (CMDS). The partitioning procedure aims at moving towards a low-dimensional feature space; the CMDS algorithm is instead exploited to set the coordinates in the mentioned low-dimensional space, and define damage indices through norms of the said coordinates. The proposed approach is shown to efficiently and robustly address the challenges linked to high-dimensional datasets and environmental variability. Results related to two large-scale test cases are reported: the ASCE structure, and the Z24 bridge. A high sensitivity to damage and a limited (if any) number of false alarms and false detections are reported, testifying the efficacy of the proposed data-driven approach.

Manifold learning based data-driven modeling for soft biological tissues.

Data-driven modeling directly utilizes experimental data with machine learning techniques to predict a material's response without the necessity of using phenomenological constitutive models. Although data-driven modeling presents a promising new approach, it has yet to be extended to the modeling of large-deformation biological tissues. Herein, we extend our recent local convexity data-driven (LCDD) framework (He and Chen, 2020) to model the mechanical response of a porcine heart mitral valve posterior leaflet. The predictability of the LCDD framework by using various combinations of biaxial and pure shear training protocols are investigated, and its effectiveness is compared with a full structural, phenomenological model modified from Zhang et al. (2016) and a continuum phenomenological Fung-type model (Tong and Fung, 1976). We show that the predictivity of the proposed LCDD nonlinear solver is generally less sensitive to the type of loading protocols (biaxial and pure shear) used in the data set, while more sensitive to the insufficient coverage of the experimental data when compared to the predictivity of the two selected phenomenological models. While no pre-defined functional form in the material model is necessary in LCDD, this study reinstates the importance of having sufficiently rich data coverage in the date-driven and machine learning type of approaches. It is also shown that the proposed LCDD method is an enhancement over the earlier distance-minimization data-driven (DMDD) against noisy data. This study demonstrates that when sufficient data is available, data-driven computing can be an alternative method for modeling complex biological materials.

Microscale Structure to Property Prediction for Additively Manufactured IN625 through Advanced Material Model Parameter Identification.

Challenge 4 of the Air Force Research Laboratory additive manufacturing modeling challenge series asks the participants to predict the grain-average elastic strain tensors of a few specific challenge grains during tensile loading, based on experimental data and extensive characterization of an IN625 test specimen. In this article, we present our strategy and computational methods for tackling this problem. During the competition stage, a characterized microstructural image from the experiment was directly used to predict the mechanical responses of certain challenge grains with a genetic algorithm-based material model identification method. Later, in the post-competition stage, a proper generalized decomposition (PGD)-based reduced order method is introduced for improved material model calibration. This data-driven reduced order method is efficient and can be used to identify complex material model parameters in the broad field of mechanics and materials science. The results in terms of absolute error have been reported for the original prediction and re-calibrated material model. The predictions show that the overall method is capable of handling large-scale computational problems for local response identification. The re-calibrated results and speed-up show promise for using PGD for material model calibration.

Data-Driven Methods for the Estimation of Leaf Water and Dry Matter Content: Performances, Potential and Limitations.

Leaf equivalent water thickness (EWT) and dry matter content (expressed as leaf mass per area (LMA)) are two critical traits for vegetation function monitoring, crop yield estimation, and precise agriculture management. Data-driven methods are widely used for remote sensing of leaf EWT and LMA because of their simplicity, satisfactory accuracy, and computation efficiency, such as the vegetation indices (VI)-based and machine learning (ML)-based methods. However, most of the data-driven methods are utilized at the canopy level, comparison of the performances of the data-driven methods at the leaf level has not been well documented. Moreover, the ML-based data-driven methods generally adopt leaf optical properties directly as their inputs, which may subsequently decrease their ability in remote sensing of leaf biochemical constituents. Performances of the ML-based methods cooperating with VI are rarely evaluated. Using the independent LOPEX and ANGERS datasets, we compared the performances of three data-driven methods: VI-based, ML-reflectance-based, and ML-VI-based methods, for the estimation of leaf EWT and LMA. Three sampling strategies were also utilized for evaluation of the generalization of these data-driven methods. Our results evidenced that ML-VI-based methods were the most accurate among these data-driven methods. Compared to the ML-reflectance-based and VI-based methods, the ML-VI-based model with support vector regression overall reduced errors by 5.7% (41.5%) and 1.8% (12.4%) for the estimation of leaf EWT (LMA), respectively. The ML-VI-based model inherits advantages of vegetation indices and ML techniques, which made it sensitive to changes of leaf biochemical constituents and capable of solving nonlinear tasks. It is thus recommended for the estimation of EWT and LMA at the leaf level. Moreover, its performance can further be enhanced by improving its generalization ability, such as adopting techniques on the selection of better wavelengths and definition of new vegetation indices. These results thus provided a prior knowledge of the data-driven methods and can be helpful for future studies on the remote sensing of leaf biochemical constituents.

Data-driven discovery of emergent behaviors in collective dynamics.

Particle- and agent-based systems are a ubiquitous modeling tool in many disciplines. We consider the fundamental problem of inferring the governing structure, i.e. interaction kernels, in a nonparametric fashion, from observations of agent-based dynamical systems. In particular, we are interested in collective dynamical systems exhibiting emergent behaviors with complicated interaction kernels, and for kernels which are parameterized by a single unknown parameter. This work extends the estimators introduced in Lu et al. (2019), which are based on suitably regularized least squares estimators, to these larger classes of systems. We provide extensive numerical evidence that the estimators provide faithful approximations to the interaction kernels, and provide accurate predictions for trajectories started at new initial conditions, both throughout the "training" time interval in which the observations were made, and often much beyond. We demonstrate these features on prototypical systems displaying collective behaviors, ranging from opinion dynamics, flocking dynamics, self-propelling particle dynamics, synchronized oscillator dynamics, to a gravitational system. Our experiments also suggest that our estimated systems can display the same emergent behaviors as the observed systems, including those that occur at larger timescales than those in the training data. Finally, in the case of families of systems governed by a parametric family of interaction kernels, we introduce novel estimators that estimate the parametric family of kernels, splitting it into a common interaction kernel and the action of parameters. We demonstrate this in the case of gravity, by learning both the "common component" 1/r 2 and the dependency on mass, without any a priori knowledge of either one, from observations of planetary motions in our solar system.

Data-Driven Method for Predicting Remaining Useful Life of Bearing Based on Bayesian Theory.

Bearings are some of the most critical industrial parts and are widely used in various types of mechanical equipment. Bearing health status can have a significant impact on the overall equipment performance, and bearing failures often cause serious economic losses and even casualties. Thus, estimating the remaining useful life (RUL) of bearings in real time is of utmost importance. This paper proposes a data-driven RUL prediction method for bearings based on Bayesian theory. First, time-domain features are extracted from the bearing vibration signal and data are fused to build a health indicator (HI) and a state model of bearing degradation. Then, according to Bayesian theory, a Bayesian model of state parameters and bearing life is established. The parameters of the Bayesian model are updated and bearing RUL is predicted by the Metropolis-Hastings algorithm. The method was validated by the XJTU-SY bearing open datasets and the prediction results are compared with the existing methods. Accuracy of the proposed method was demonstrated.