Low Error Research Articles

The joint effect of changes in urbanization and climate on trends in floods: A comparison of panel and single-station quantile regression approaches

Estimates of annual maximum (peak) flow quantiles are needed for basins undergoing changes in both urbanization and climate. Most previous work on the effect of urbanization on peak flows has considered urbanization alone and only the spatial variation in flood quantiles or its mean temporal effect, and most work on the effect of nonstationarity in climate has focused on single-station analyses, which give uncertain results for extreme quantiles. To address these gaps, three approaches to the statistical estimation of the joint effects of changes in impervious cover and climate on the estimation of peak-flow quantiles were compared: single-station quantile regression; a fixed effect panel-quantile regression (pQR) method using a location (mean) shift to homogenize the panel; and a location-scale panel regression model (pQRmom), which accounts for both scale (variance) and location effects. The different approaches were applied to a dataset consisting of instantaneous annual peak flows from 127 minimally nested basins in the midwestern United States with at least 4 % change in imperviousness. The annual maximum daily discharge from a water-balance model was selected as the primary climate predictor; in addition, to provide a comparison of climate predictors, precipitation was also considered. The coefficients from single-station regressions were usually sufficiently certain to determine the effects of climate variation but usually too uncertain to estimate the effects of urbanization. The panel-quantile regression approaches give much more certain results, but their estimates of quantile dependence differ: although both indicate urbanization effects decreasing with decreasing annual exceedance probability (AEP), the pQRmom urbanization coefficients are insignificantly different from zero for AEPs less than 0.10, whereas the pQR coefficients remain positive and are significant except for AEP = 0.01, the smallest AEP value considered. Although the location-scale structure of the pQRmom approach has less flexible quantile dependence than the pQR approach, the pQRmom approach has somewhat lower overall error, and it is found that by subsetting the dataset to homogenize the scale effects, the pQR and pQRmom results become similar, indicating the insignificant urbanization coefficients for small AEPs of the pQRmom results are likely correct for the study dataset.

IGGCN: Individual-Guided Graph Convolution Network for Pedestrian Trajectory Prediction

Accurately predicting the future trajectory of pedestrians is crucial for applications such as autonomous driving and robot navigation. Graph convolution is widely used in trajectory prediction tasks due to its scalability and adaptive feature-learning capabilities. However, there are two problems with pedestrian trajectory prediction methods based on graph convolution: 1. Previous methods struggled to adjust social interactions according to the attributes of different pedestrians, making it difficult to accurately model the relative importance between different pedestrians and others; 2. Previous methods lacked dynamic processing of pedestrian spatial-temporal interaction features to capture high-level spatial-temporal interaction features effectively. Therefore, we propose an Individual-Guided Graph Convolution Network (IGGCN) for pedestrian trajectory prediction. To tackle problem 1, we design an individual-guided interaction module that can adjust pedestrian social interaction modeling according to the pedestrian's attributes, thereby achieving an accurate description of the relative importance of pedestrians. We extend the module to temporal interaction modeling to further achieve an accurate description of the relative importance of time frames. To address problem 2, we design a deformable convolution module to dynamically process spatial-temporal interaction features through deformable convolution kernels, facilitating the capture of high-level spatial-temporal interaction features. We evaluate our method on the ETH, UCY, and SDD datasets. Quantitative analysis shows that our method has lower prediction errors than the current state-of-the-art methods. Qualitative analysis further reveals that our method effectively eliminates the influence of irrelevant pedestrians and accurately models the spatial-temporal interaction relationship of pedestrians.

Effect of Chiropractic Intervention on Oculomotor and Attentional Visual Outcomes in Young Adults With Long-Term Mild Traumatic Brain Injury: A Randomized Controlled Trial

ObjectiveThis study aimed to establish if chiropractic care can improve oculomotor and cognitive symptoms in individuals with persistent postconcussion syndrome (PPCS). MethodsA single-blind, randomized controlled intervention study recorded baseline computerized eye-tracker assessment (CEA) outcomes in 40 young adults with PPCS following mild traumatic brain injury. Participants were randomly allocated to either a chiropractic or age-matched active control intervention, and the change in CEA outcomes following intervention was compared between the chiropractic and control groups. A battery of CEAs including egocentric localization, fixation stability, pursuit, saccades, Stroop, and the vestibulo-ocular reflex, were used to assess oculomotor function, visual attention/processing, and selective attention. ResultsRelative to the control group, participants receiving the chiropractic intervention scored better in the Stroop test (P < .001), had improved gaze stability during both vestibulo-ocular reflex (P < .001) and fixation stability (P = .009), and a lower vertical error in egocentric localization (P < .001). However, performance was poorer in pursuits, where they had an increased tracking error (P < .001). ConclusionChiropractic care in participants with PPCS significantly improved static and dynamic gaze stability, and performance in the Stroop test, compared with a control intervention. These results suggest that chiropractic care can offer a novel avenue for alleviating certain visual and cognitive symptoms in patients with PPCS. It also adds to the growing evidence that suggests that some longstanding PPCS visual symptoms may have a spinal or proprioceptive basis.

Assessing forecast performance of daily reference evapotranspiration: A comparison of equations, machine and deep learning using weather forecasts

An accurate forecast of short-term reference evapotranspiration (ET0) is crucial for effective farm and irrigation scheduling management. The techniques of forecasting ET0 have progressed from empirical equations to artificial intelligence-based models. However, gaps have remained in comprehensively evaluating the ET0 forecast performance discrepancies between various methods based on weather forecasts in addition to historical meteorological data, as well as investigating the effect of the amount of input variables on ET0 forecast accuracy. Here, the study evaluated the forecast accuracy of five conventional equations calibrated and six machine learning (ML) and deep learning (DL) models in predicting daily ET0 for a lead time of 1–7 days in the North China Plain (NCP). Comparative evaluations indicate that the Analytical Penman-Monteith (APM) is the optimum empirical equation for forecasting ET0, but its accuracy decreases as lead times increase. Conversely, ML and DL models consistently outperform empirical equations, characterized by lower mean absolute error (MAE) and root mean square error (RMSE) as well as correlation coefficient and Nash-Sutcliffe efficiency coefficient approaching 1 across all lead times. Notably, the Bidirectional Long short-term memory (Bi-LSTM) model outperforms other models, maintaining robust and effective forecast performance even on day 7. Moreover, with the increasing input variables, combining weather forecasts and historical meteorological data, the forecast performance of all models is improved, with significant enhancement observed in the Bi-LSTM model. This work provides valuable insights for choosing appropriate models to forecast ET0, holding essential in regional managing farm and irrigation systems.

Data-enhanced convolutional network based on air conditioning system start/stop time prediction

Most enterprise workshop operators frequently adjust the start/stop time of air conditioning systems based on indoor and outdoor temperatures and humidity to accommodate changing demand and weather conditions. However, relying on personal subjective experience for these adjustments often leads to operational delays or energy waste due to the lack of precision in determining optimal timing. Predicting air conditioning system start and stop times is crucial for energy consumption and savings in HVAC systems. Traditional data-driven methods have been insufficient in this regard, as they mainly focus on feature mapping and overlook the dynamic coupling relationships of process variables, resulting in subpar predictions. In response to this challenge, the paper introduces a novel approach known as the Periodicity and Long-Term Convolutional Neural Network (PLCNN). This method converts one-dimensional regression prediction data into two-dimensional data containing time series features to capture the dynamic coupling characteristics of the air conditioning system while maintaining the independent variation relationships of features. Experimental results using real factory floor data have demonstrated the superior performance of the PLCNN method. Specifically, this method achieved a 14.96% lower error rate compared to the traditional method and an 8.18% improvement compared to the deep learning method. Moreover, the implementation of the PLCNN method in the optimal control of air conditioning systems led to a significant 19.43% reduction in total monthly energy consumption. In conclusion, the proposed method offers a promising alternative to traditional approaches to forecasting and provides a solution to the common challenges encountered in traditional prediction tasks.

Segmented Two-Dimensional Progressive Polynomial Calibration Method for Nonlinear Sensors.

Nonlinearity in sensor measurements reduces the sensor's accuracy. Therefore, accurate calibration is necessary for reliable sensor operation. This study proposes a segmented calibration method that divides the input range into multiple sections and calculates the optimized calibration functions for each one. This approach reduces the overall error rate and improves the calibration accuracy by isolating distinctive regions. The modified progressive polynomial calibration technique is used to calculate the calibration function. This algorithm addresses the computational complexity, allowing for reduced polynomial degrees and improving the accuracy. The segmented calibration method achieves a significantly lower error rate of 0.000006% compared to the original single calibration method, which has an error rate of 0.0823%, when using the same six calibration points and a fifth-degree polynomial function. This method maintains improved accuracy with fewer calibration points, and its ability to reduce the computational complexity and calculation time while using lower polynomial degrees is confirmed. Additionally, it can be extended to two dimensions to reduce the errors caused by cross-sensitivity. The results from a two-dimensional simulation show a reduction in the error rate ranging from 15.84% to 2.07% in an 8-bit signed fixed-point system. These results indicate that the segmented calibration method is an effective and scalable solution for various typical sensors.

GNSS-R-Based wildfire detection: a novel and accurate method

ABSTRACT In recent years, global forests have faced frequent wildfires due to climate change, leading to significant ecological and economic losses. Traditional fire monitoring technologies like MODIS and VIIRS are often hindered by cloud cover and smoke. This study explores the potential of Global Navigation Satellite System Reflectometry (GNSS-R) for fire monitoring in complex environments. By analyzing cumulative burned areas in southeastern Brazil, a new fire detection method combining GNSS-R with Convolutional Neural Networks (CNN) is proposed to improve accuracy and response time. Results show high agreement between predicted and actual fire areas, with CNN outperforming Random Forest (RF) by 3.3% in prediction accuracy and achieving a 17.4% increase in model fitting. Error analysis indicates no systematic errors, with CNN exhibiting lower fire detection errors and better spatial alignment with actual conditions than RF, which tends to overestimate. This study provides a novel approach to fire detection using GNSS-R, with significant implications for fire prediction and management strategies.

A novel recursive sub-tensor hyperspectral compressive sensing of plant leaves based on multiple arbitrary-shape regions of interest

Plant hyperspectral images (HSIs) contain valuable information for agricultural disaster prediction, biomass estimation, and other applications. However, they also include a lot of irrelevant background information, which wastes storage resources. In this paper, we propose a novel recursive sub-tensor hyperspectral compressive sensing method for plant leaves. This method uses recursive sub-tensor compressive sensing to compress and reconstruct each arbitrary-shape leaf region, discarding a large amount of background information to achieve the best possible reconstruction performance of the leaf region and significantly reduce storage space. The proposed method involves several key steps. Firstly, the optimal band is determined using the spatial spectral decorrelation criterion, and its corresponding mask image is used to extract the leaf regions from the background. Secondly, the recursive maximum inscribed rectangle algorithm is applied to obtain rectangular sub-tensors of leaves recursively. Each sub-tensor is then individually compressed and reconstructed. Finally, all sub-tensors can be reconstructed to form complete leaf HSIs without background information. Experimental results demonstrate that the proposed method achieves superior image reconstruction quality at extremely low sampling rates compared to other methods. The proposed method can improve average Peak Signal-to-Noise Ratio (PSNR) values by about 3.04% and 0.74% compared to Tensor Compressive Sensing (TCS) at the sampling rate of 2%. In the spectral domain, the proposed method can achieve significantly smaller Spectral Angle Mapper (SAM) values and relatively lower spectral indices errors for Double Difference, Triangular Vegetation Index, Leaf Chlorophyll Index, and Modified Normalized Difference 680 than those of TCS. Therefore, the proposed method achieves better compression performance for reconstructed plant leaf HSIs than the other methods.

Spatiotemporal wind speed forecasting using conditional local convolution and multidimensional meteorology features

Wind speed prediction is crucial for precisely wind power forecasting and reduced maintenance costs. Highland regions, which possess a considerable wind potential, present complex meteorological conditions, making wind speed prediction challenging. Traditional weather forecasting relies on complex statistical methods and extensive prior knowledge. While recent deep learning models have improved prediction accuracy, they often assume uniform influence weight structure, limiting model effectiveness. This study introduces an enhanced Conditional Local Convolution Recurrent Network (CLCRN) model to improve spatiotemporal wind speed forecasting using multidimensional meteorological inputs such as temperature, pressure, and dew point, alongside wind components. This model addresses uniform influence model weight issue by redesigning convolution kernels to better capture local meteorological features and integrating multiple influencing factors. Our model consistently achieves lower Mean Absolute Error (MAE) and Root Mean Squared Error (RMSE) values across various prediction intervals (3, 6, 9, and 12 h) compared to other models, supported by the meteorological station data from 2019 to 2021. Furthermore, the spatial distribution of the local convolution weights aligns with local wind velocity patterns in Inner Mongolia, enhancing model interpretability. These results demonstrate potential for practical applications in renewable energy planning and wind dynamics simulation.

Intensive short-term sampling with long-term consequences: characterizing pollutant transport with implications for developing monitoring.

Riverine sampling of pollutants is commonly used to understand pollutants' transport pathways, relationships with hydrology, and overall presence in a waterbody. However, temporal gaps between sample collection introduce errors to these efforts, and guidance prescribing sampling frequency remains sparse. The magnitude of error often depends on a contaminant's transport mechanisms and local hydrologic conditions, making the creation of comprehensive sampling guidance difficult. This study analyzed a unique dataset that measured 18 analytes, including pesticides, nutrients, and pathogens, in three Iowa rivers for 90 consecutive days (May 4-August 1, 2000). This dataset provided a novel opportunity to relate pollutants to local hydrology and quantify errors associated with recurring sampling. Pesticide concentrations followed the spring flush phenomenon, where values were greatest during high streamflow in May and June but often depleted by July. Fecal coliform and total phosphorus (TP) also coincided with high flow, but unlike pesticides, their concentrations never diminished. Nitrate exhibited more complex behavior; concentrations were diluted during high flows and then increased as streamflow receded. Autocorrelations were significant for nitrate and atrazine in larger rivers but negligible for fecal coliform and TP. Loads were calculated for four pollutants with minimal non-detects (atrazine, fecal coliform, nitrate, and TP). We simulated intermittent sampling by selecting evenly spaced subsets of measured values to estimate loads, which were compared to the loads calculated using every daily sample to quantify error. This method typically overestimated nitrate loads but underestimated other pollutants, and errors often decreased in larger watersheds. Nitrate generally had the lowest error, while fecal coliform had the highest. We used these results to approximate the sampling frequency needed to bind errors within a certain threshold.

Integrated metaheuristic approaches for estimation of fracture porosity derived from fullbore formation micro-imager logs: Reaping the benefits of stand-alone and ensemble machine learning models

Fracture porosity is one of the most effective parameters for reservoir productivity and recovery efficiency. This study aims to predict and improve the accuracy of fracture porosity estimation through the application of advanced machine learning (ML) algorithms. A novel approach was introduced for the first time to estimate fracture porosity by reaping the benefits of petrophysical and fullbore formation micro-imager (FMI) data based on employing various stand-alone, ensemble, optimisation and multi-variable linear regression (MVLR) algorithms. This study proposes a ground-breaking two-step committee machine (CM) model. Petrophysical data containing compressional sonic-log travel time, deep resistivity, neutron porosity and bulk density (inputs), along with FMI-derived fracture porosity values (outputs), were employed. Nine stand-alone ML algorithms, including back-propagation neural network, Takagi and Sugeno fuzzy system, adaptive neuro-fuzzy inference system, decision tree, radial basis function, extreme gradient boosting, least-squares boosting, least squares support vector regression and k-nearest neighbours, were trained for initial estimation. To improve the efficacy of stand-alone algorithms, their outputs were combined in CM structures using optimisation algorithms. This integration was applied through five optimisation algorithms, including genetic algorithm, ant colony, particle swarm, covariance matrix adaptation evolution strategy (CMA-ES) and Coyote optimisation algorithm. Considering the lowest error, the CM with CMA-ES showed superior performance. Subsequently, MVLR was employed to improve the CMs further. Employing MVLR to combine the CMs yielded a 57.85% decline in mean squared error and a 4.502% improvement in the correlation coefficient compared to the stand-alone algorithms. The results of the benchmark analysis validated the efficacy of this approach.

Enhancing the MCL-SLAM algorithm to overcome the issue of illumination variation, non-static environment and kidnapping to present the NIK-SLAM

AbstractSimultaneous Localization and Mapping (SLAM) is a major challenge that has been comprehensively investigated by scholars. SLAM’s ability to support autonomous navigation helps the robot to move from one place to another without being controlled by humans. This achievement has attracted the attention of scholars towards focusing on this area. Over the decades, countless procedures have been developed with extraordinary accomplishments. However, some problems can degrade the efficiency of SLAM (Simultaneous Localization and Mapping) technique, and examples of such problems include environmental noises (light intensity and shadow), non-static environment, and kidnap robots. These problems create inconsistency in measurement that can lead to erroneous navigation for an unmanned vehicle. In minimizing the effect of these problems, a novel SLAM technique based on several re-modifications of MCL is presented in this paper. This innovative technique is known as NIK-SLAM. The NIK-SLAM is equipped with filters alongside some modifications of the original Monte-Carlo algorithm (MCL) to maximize its performance. The filters were employed to overcome the effect of shadow and light intensity while the re-modification of MCL is used to overcome the limitation of kidnapping and non-static objects. A publicly available dataset (TUM-RGBD) and a private dataset were used for evaluation. Matlab was used for simulation while assessments were based on quantitative and qualitative validation. The results demonstrated a better performance of the NIK-SLAM, attaining a lower trajectory error for most experiments when compared to the original MCL and other SLAM algorithms found in the literature, but at the expense of delayed processing time. Furthermore, this accomplishment supports pathway planning, and exploration into unstable environments while it decreases accident rates/human fatalities, and in the future study, the issue of delayed processing time will be addressed.

Application of Reservoir Computers for Chaos Synchronization and Cracking Chaos-Based Cryptography in Fermionic 2D Thirring and 4D Gursey Models with Spinor Fields

Reservoir computers have recently become one of the most widely exploited model-free approaches to chaos synchronization due to their fast convergence speed and simple yet effective learning rules. In this study, reservoir computing has been utilized to emulate the dynamics of chaotic Thirring and Gursey systems. In the supervised learning of the reservoir, one-step ahead states of the dynamical systems have been employed as the teaching signals. After the training phase, the reservoir was first run autonomously and then weakly driven by chaotic systems. It has been shown that the trained reservoir computers can exhibit the same characteristics as the attractors of the learned chaotic systems, which enable their use in synchronization tasks of chaotic systems with possible application in cracking chaos-based cryptography. To investigate the effect of noise on the performance of reservoir computers, noise signals with different colors and amplitudes have been included in the transmitted signals. The obtained results indicate that the proposed scheme can provide lower error metrics for both models.

Multi-task Learning for Gaussian Graphical Regressions with High Dimensional Covariates

Gaussian graphical regression is a powerful approach for regressing the precision matrix of a Gaussian graphical model on covariates, which permits the response variables and covariates to outnumber the sample size. However, traditional approaches of fitting the model via separate node-wise lasso regressions overlook the network-induced structure among these regressions, leading to high error rates, particularly when the number of nodes is large. To address this issue, we propose a multi-task learning estimator for fitting Gaussian graphical regression models, which incorporates a cross-task group sparsity penalty and a within-task element-wise sparsity penalty to govern the sparsity of active covariates and their effects on the graph, respectively. We also develop an efficient augmented Lagrangian algorithm for computation, which solves subproblems with a semi-smooth Newton method. We further prove that our multi-task learning estimator has considerably lower error rates than the separate node-wise regression estimates, as the cross-task penalty enables borrowing information across tasks. We examine the utility of our method through simulations and an application to a gene co-expression network study with brain cancer patients.

Assessing the Predictive Capabilities of Autoregressive Integrated Moving Average and Linear Regression Models for Acute Changes in Clinical and Selected Laboratory Parameters in Children After Cardiac Surgery in the ICU

(1) Background: The main objective of this research was to assess the clinical factors related to the condition of pediatric patients with congenital heart defects after they underwent intensive care unit surgery. The information was gathered from the Congenital Heart Disease Surgery Unit at the National Heart Foundation Hospital and Research Institute in Dhaka, Bangladesh. We gathered and examined data from 288 ICU patients. Patients under the age of twelve who required more than a 24-h ICU stay were selected. (2) Methods: The dependent and independent variables were chosen in advance based on expert opinion. The relationships between these pre-specified ICU parameters were determined using the Pearson correlation model and assessed through linear regression and ARIMA modeling to predict subsequent acute changes in the patients’ ICU statuses. (3) Results: A statistically significant relationship (p value &lt; 0.001) was found between CVP and BP (95% CI = 0.2113; 0.353 r = 0.2841249) and between PEEP and FiO2 (95% CI = 0.6992; 0.770 r = 0.7367744). Although the relationships between pH and PO2 were minor (95% CI = 0.161; 0.308 r = 0.2362575), they were statistically significant. The parameters considered statistically significant (p &lt; 0.001) were chosen for forecasting. In this work, the linear regression model and the ARIMA model used the parameters BP, FiO2, and PO2 for prediction. We forecasted the patients’ statuses for the next hour. It was found that the ARIMA model had a lower error rate than the linear regression model. (4) Conclusions: This study helps identify the important parameters for predicting and monitoring patients’ statuses in the ICU, with the ultimate goal of providing physicians with an early warning system to anticipate deterioration in clinical and biochemical parameters. The ability to accurately forecast future patients’ conditions can enable proactive, targeted interventions, potentially improving outcomes and reducing the risk of adverse events.

Time series forecasting of rain-induced attenuation using coupled ARIMA-SARIMA models for radio propagation applications over subtropical locations

This present research aimed to identify the optimum model for predicting the rain-induced attenuation along the terrestrial and satellite communication networks and employed the time series forecasting models of Autoregressive integrated moving average (ARIMA) and Seasonal Autoregressive integrated moving average (SARIMA) to examine and contrast several techniques for predicting rain-induced attenuation across a sub-tropical area of South Africa. In this research, the ARIMA was developed using the Box and Jenkins technique to predict long-term rain-induced attenuation in the following South African provinces: Kwazulu-Natal, Eastern Cape, Gauteng, and Northern Cape. The datasets used for the research were obtained from the South African Weather Station from 1994–2023 were used to build and check the model after generating rain-induced attenuation using synthetic storm techniques (SST). The forecasting performance of the seasonal autoregressive integrated moving average (SARIMA) model and that of autoregressive integrated moving average (ARIMA) were compared with four forecast performance measures: - Mean Squared Error (MSE), Mean Absolute Error (MAE), R- squared or the coefficient of determination (R2) and Root Mean Squared Error (RMSE). The results of the study showed that due to its lower forecast performance error indicators, SARIMA performed better than ARIMA in forecasting rain-induced attenuation in South Africa. There is no statistically significant difference between the two projected values, according to the results of a Ljung-box test for significant difference. The authors draw the inference that the two approaches can effectively be employed in their proper locations. Rain-attenuation prediction data indicates that this long-term projection can help decision makers to improve the performance of microwave networks in the face of random fluctuations and avoiding unexpected signal loss with efficient installation of communications infrastructure. It also has applications in prediction of floods, urban planning, and crop management.

Incorporating crumb rubber in slag-based geopolymer: Experimental work and predictive modelling

Crumb rubber (CR), a prevalent hazardous waste material, poses significant environmental challenges that necessitate innovative management strategies. Recent studies have focused on incorporating CR and ground granulated blast furnace slag (GGBS) into geopolymer concrete (GC) to address this challenge. This research presents a novel approach that combines experimental analysis with machine learning (ML) techniques to investigate the chemical effects of these materials on GC. Initially, compressive strength (CS) tests are conducted on rubberized geopolymer (RG) concrete, and the resulting data are further analyzed through ML algorithms to enhance the understanding of material behavior. Despite numerous attempts by researchers to integrate CR into geopolymer-based materials, the challenge of mitigating strength degradation persists. Therefore, it becomes imperative to construct a predictive model that relies on new variables influencing the strength characteristics. This research develops a predictive model to assess concrete compressive strength (CS). Key variables such as CR grade (particle size), incorporation percentage, and oxide ratio were analyzed for their impact on geopolymer strength. Employing ensemble techniques, namely Bagging (BA) and Gradient boosting (GB) with five learners (M5P, random forest (RF), random tree (RT), reduced error pruning tree (REPT), and support vector machine (SVM), the study found that M5P-GB models offered the most accurate CS predictions with the highest accuracy and lowest errors. Moreover, the dataset was checked via the k-fold cross-validation technique and confirmed the developed models' robustness, reliability, and effectiveness. Furthermore, uncertainty analysis revealed that GB-M5P-unpruned models-maintained uncertainty percentages of 14.769 % at 7 days and 11.277 % at 28 days, which was under the 35 % threshold. Additionally, the sensitivity analysis identified the CR grade and replacement percentage by volume of crusher dust (CD) as the most critical factors affecting CS. These findings underscore the importance of considering predictive model development in decision-making processes related to RG concrete properties, providing valuable insights into optimizing the model's robustness and reliability for practical applications. However, the study's reliance on specific input parameters and controlled laboratory conditions may limit the generalizability of the predictive models, necessitating further validation under diverse real-world conditions and variations in CR source and environmental factors.

Soil and Water Assessment Tool (SWAT)-Informed Deep Learning for Streamflow Forecasting with Remote Sensing and In Situ Precipitation and Discharge Observations

In order to anticipate residual errors and improve accuracy while reducing uncertainties, this work integrates the long short-term memory (LSTM) with the Soil and Water Assessment Tool (SWAT) to create a deep learning (DL) model that is guided by physics. By forecasting the residual errors of the SWAT model, the SWAT-informed LSTM model (LSTM-SWAT) differs from typical LSTM approaches that predict the streamflow directly. Through numerical tests, the performance of the LSTM-SWAT was evaluated with both LSTM-only and SWAT-only models in the Upper Heihe River Basin. The outcomes showed that the LSTM-SWAT performed better than the other models, showing higher accuracy and a lower mean absolute error (MAE = 3.13 m3/s). Sensitivity experiments further showed how the quality of the training dataset affects the performance of the LSTM-SWAT. The results of this study demonstrate how the LSTM-SWAT may improve streamflow prediction greatly by remote sensing and in situ observations. Additionally, this study emphasizes the need for detailed consideration of specific sources of uncertainty to further improve the predictive capabilities of the hybrid model.

Significant of Blood Glucose Time-Series Forecasting with Temporal Fusion Transformer Model for Type 1 Diabetes

Blood glucose (BG) prediction has advanced to a state of the art thanks to deep learning models, which have been demonstrated to enhance type 1 diabetes (T1D) therapy. Nevertheless, most current models are limited to single-horizon prediction and have several practical drawbacks, including poor interpretability. In this study, we develop a novel approach to optimize the forecasting model in the training processes using the Optuna function, cross-validation, and the latest dataset. We offer a novel deep learning framework for multi-horizon BG prediction: the temporal fusion transformer (TFT). TFT employs a self-attention mechanism to extract long-term temporal dependencies and enables a model with an auto-tuning adjustment approach on hyperparameters and a cross-validation function on univariate and multivariate input models. On a clinical dataset with T1D subjects, namely ShanghaiT1DM (16 subjects), D1NAMO (9 subjects), and OhioT1DM (6 subjects) datasets, it achieved an average root mean square error (RMSE) of the three datasets used. The TFT model gave a lower error score than the baseline models of neural hierarchical interpolation for time series (N-HiTS)and long short-term memory (LSTM) with the values of 10.08+0.31 mg/dL and 12.34+0.62 mg/dL on the univariate input model and 9.18+1.21 mg/dL and 14.33+0.52 mg/dL on the multivariate input model for the prediction horizons of 30 and 60 minutes, respectively. This result explained that the TFT model was adequate for carrying out multi-horizon BG value-level forecasting and potentially deploying on-edge devices to improve clinical efforts to manage BG levels for T1D patients in real-time application.

Neural fields for rapid aircraft aerodynamics simulations

This paper presents a methodology to learn surrogate models of steady state fluid dynamics simulations on meshed domains, based on Implicit Neural Representations (INRs). The proposed models can be applied directly to unstructured domains for different flow conditions, handle non-parametric 3D geometric variations, and generalize to unseen shapes at test time. The coordinate-based formulation naturally leads to robustness with respect to discretization, allowing an excellent trade-off between computational cost (memory footprint and training time) and accuracy. The method is demonstrated on two industrially relevant applications: a RANS dataset of the two-dimensional compressible flow over a transonic airfoil and a dataset of the surface pressure distribution over 3D wings, including shape, inflow condition, and control surface deflection variations. On the considered test cases, our approach achieves a more than three times lower test error and significantly improves generalization error on unseen geometries compared to state-of-the-art Graph Neural Network architectures. Remarkably, the method can drastically accelerate high fidelity numerical solver, achieving a five order of magnitude speedup on the RANS transonic airfoil dataset.