Search Results (30,992)

When complex computations cannot be performed on board a mobile robot, sensory data must be transmitted to a remote station to be processed, and the resulting actions must be sent back to the robot to execute, forming a repeating cycle. This involves stochastic round-trip times in the case of non-deterministic network communications and/or non-hard real-time software. Since robots need to react within strict time constraints, modeling these round-trip times becomes essential for many tasks. Modern approaches for modeling sequences of data are mostly based on time-series forecasting techniques, which impose a computational cost that may be prohibitive for real-time operation, do not consider all the delay sources existing in the sw/hw system, or do not work fully online, i.e., within the time of the current round-trip. Marginal probabilistic models, on the other hand, often have a lower cost, since they discard temporal dependencies between successive measurements of round-trip times, a suitable approximation when regime changes are properly handled given the typically stationary nature of these round-trip times. In this paper we focus on the hypothesis tests needed for marginal modeling of the round-trip times in remotely operated robotic systems with the presence of abrupt changes in regimes. We analyze in depth three common models, namely Log-logistic, Log-normal, and Exponential, and propose some modifications of parameter estimators for them and new thresholds for well-known goodness-of-fit tests, which are aimed at the particularities of our setting. We then evaluate our proposal on a dataset gathered from a variety of networked robot scenarios, both real and simulated; through >2100 h of high-performance computer processing, we assess the statistical robustness and practical suitability of these methods for these kinds of robotic applications. Full article

Ultra-high-performance concrete (UHPC) is increasingly employed in long-span and heavily loaded structural applications; however, the accurate prediction of its flexural capacity remains a significant challenge because of the complex interactions among geometric parameters, reinforcement details, and advanced material properties. Existing design codes and single-architecture machine learning models often struggle to capture these nonlinear relationships, particularly when experimental datasets are limited in size and diversity. This study proposes a compact hybrid CNN–Transformer model that combines convolutional layers for local feature extraction with self-attention mechanisms for modeling long-range dependencies, enabling robust learning from a database of 120 UHPC beam tests drawn from 13 laboratories worldwide. The model’s predictive performance is benchmarked against conventional design codes, analytical and semi-empirical formulations, and alternative machine learning approaches including Convolutional Neural Networks (CNN), eXtreme Gradient Boosting (XGBoost), and K-Nearest Neighbors (KNN). Results show that the proposed architecture achieves the highest accuracy with an R² of 0.943, an RMSE of 41.310, and a 25% reduction in RMSE compared with the best-performing baseline, while maintaining strong generalization across varying fiber dosages, reinforcement ratios, and shear-span ratios. Model interpretation via SHapley Additive exPlanations (SHAP) analysis identifies key parameters influencing capacity, providing actionable design insights. The findings demonstrate the potential of hybrid deep-learning frameworks to improve structural performance prediction for UHPC beams and lay the groundwork for future integration into reliability-based design codes. Full article

Accurate prediction of shear strength at the interface between new and old concrete is vital for the structural performance of repaired and composite systems. However, the underlying shear transfer mechanism is highly nonlinear and influenced by multiple interdependent factors, which limit the applicability of conventional empirical models. To address this challenge, an interpretable machine-learning (ML) framework is proposed. The latest database of 247 push-off specimens was compiled from the recent literature, incorporating diverse interface types and design parameters. The hyperparameters of the adopted ML models were optimized via a grid search to ensure the predictive performance on the updated database. Among the evaluated algorithms, eXtreme Gradient Boosting (XGBoost) demonstrated the best predictive performance, with R² = 0.933, RMSE = 0.663, MAE = 0.486, and MAPE = 12.937% on the testing set, outperforming Support Vector Regression (SVR), Random Forest (RF), and adaptive boosting (AdaBoost). Compared with the best empirical model (AASHTO, R² = 0.939), XGBoost achieved significantly lower prediction errors (e.g., RMSE was reduced by 67.8%), enhanced robustness (COV = 0.176 vs. 0.384), and a more balanced mean ratio (1.054 vs. 1.514). The SHapley Additive exPlanations (SHAP) method was employed to interpret the model predictions, identifying the shear reinforcement ratio as the most influential factor, followed by interface type, interface width, and concrete strength. These results confirm the superior accuracy, generalizability, and explainability of XGBoost in modeling the shear behaviors of new–old concrete interfaces. Full article

Background/Objectives: This study aims to build an interpretable and accurate predictive model for myocardial infarction (MI) using Explainable Boosting Machines (EBM), a state-of-the-art Explainable Artificial Intelligence (XAI) technique. The objective is to identify and rank clinically relevant biomarkers that contribute to MI diagnosis while maintaining transparency to support clinical decision making. Methods: The dataset comprises 1319 patient records collected in 2018 from a cardiology center in the Erbil region of Iraq. Each record includes eight routinely measured clinical and biochemical features, such as troponin, CK-MB, and glucose levels, and a binary outcome variable indicating the presence or absence of MI. After preprocessing (e.g., one-hot encoding, normalization), the EBM model was trained using 80% of the data and tested on the remaining 20%. Model performance was evaluated using standard metrics including AUC, accuracy, sensitivity, specificity, F1 score, and Matthews correlation coefficient. Feature importance was assessed to identify key predictors. Partial dependence analyses provided insights into how each variable affected model predictions. Results: The EBM model demonstrated excellent diagnostic performance, achieving an AUC of 0.980, an accuracy of 96.6%, sensitivity of 96.8%, and specificity of 96.2%. Troponin and CK-MB were identified as the top predictors, confirming their established clinical relevance in MI diagnosis. In contrast, demographic and hemodynamic variables such as age and blood pressure contributed minimally. Partial dependence plots revealed non-linear effects of key biomarkers. Local explanation plots demonstrated the model’s ability to make confident, interpretable predictions for both positive and negative cases. Conclusions: The findings highlight the potential of EBM as a clinically useful and ethical AI approach for MI diagnosis. By combining high predictive accuracy with transparency, EBM supports biomarker prioritization and clinical risk stratification, thus aligning with precision medicine and responsible AI principles. Future research should validate the model on multi-center datasets and explore additional features for broader clinical use. Full article

A pivotal food crop in arid and semi-arid zones, proso millet boasts remarkable economic value, making the breeding of stable high-yield varieties critical for industrial sustainability. This study employed a randomized complete block design to conduct a two-year multi-environment trial on nine new varieties across six representative spring-sown test regions in China. Analytical tools, including additive main effects and multiplicative interaction (AMMI) biplots, AMMI stability values (ASV), genotype and genotype × environment (GGE) models, and genotype by yield–trait (GYT) biplots were utilized to assess genotype–environment (G × E) interactions and screen superior genotypes. AMMI variance analysis showed extremely significant effects of genotype, environment, and G × E on yield traits (p < 0.01). G × E principal component analysis identified JS8, PS3, PS6, and PM4 as dominant genotypes. Based on ASV indices, varietal stability rankings were PS5 > YS13 > JS8 > PS3 > PS6 > PM4 > others. GGE analysis indicated PM4 had the broadest adaptability across tested environments, while JS15 exhibited specific adaptability in Datong. Huairen and Shuozhou were validated as ideal testing environments via an ideal environment plot. GYT biplots further confirmed that YS13, JS15, PS3, and PM4 excelled in comprehensive yield–trait combinations. These findings offer a scientific foundation for ecological adaptability evaluation, breeding material selection, and commercial variety promotion. Full article

Inhibitors of the tyrosine kinase Zap70 are actively searched to improve treatments of lymphoid malignancies and autoimmune diseases associated with an abnormal T-cell response. The natural product withaferin A (WFA) has been characterized as a covalent inhibitor of Zap70 capable of blocking the migration of human T-cells. By analogy, we postulated that other withanolides equipped with a thiol-reactive, α,β-unsaturated ketone may form covalent complexes with Zap70. The hypothesis was tested using a molecular modeling approach with a panel of 12 withanolides docked onto the kinase domain of Zap70. Seven natural products revealed a capability to form stable complexes with Zap70 comparable to that of WFA, including withangulatin A, 4β-hydroxywithanolide E, withaperuvin, and ixocarpalactone A. Withangulatin A surpassed all the other withanolides for its ability to engage an interaction with Zap70 kinase and to form covalent complexes via bonding to the Cys346 residue close to the enzyme active site. The physicochemical and ADMET properties of withangulatin A were analyzed via Density Functional Theory calculations and an analysis of its Fukui function descriptors. The C3 position of the enone moiety was identified as the most reactive (nucleophilic) site of the molecule. Withangulatin A revealed a satisfactory ADMET profile with no major toxicity anticipated. It represents a potential hit to guide the design of Zap70 inhibitors. Full article

Determining the active substance content in the tested product is an essential part of research for overall assessment of the quality of a medicinal substance. This role can be successfully performed by membrane electrodes that are selective for a specific drug. The novelty of the presented research is the development of the first ion-selective electrode with a polymer membrane phase with the octenidine (OCT) function. Classical ion-selective electrodes (ISE), polymer electrodes with an internal Ag/AgCl electrode, and electrode bodies with glassy carbon were used for the research. The membranes were prepared based on cation exchangers from the borate group and neutral cyclodextrin. All sensors have good parameters, e.g., the polymer electrode with KtpClPB is characterised by a wide linear range of −logc 6−3, a low limit of detection 5 × 10⁻⁷ M, and a near-Nernstian, reproducible slope of characteristics of 31.41 ± 1.14 mV/decade. It can be seen that a stable, reversible potential and a short response time were achieved for this sensor. The obtained favourable selectivity coefficients of the electrode determined in relation to excipients allowed direct determination of octenidine, e.g., in lozenges. The results obtained with the calibration curve method show a recovery of 97% and a precision of SD 2.3 mg/L, which indicates that the data are consistent with the pharmacopoeia requirements. Full article

Foodborne pathogenic bacteria, like Salmonella spp. and Listeria monocytogenes, can be detected in the primary food production environment. On the other hand, and in the current context of One Health, antimicrobial resistance (AMR) is gaining increased attention worldwide, as it poses significant threat to public health. The purpose of this study was to confirm the presence of Salmonella spp. and L. monocytogenes in pet food and feed samples, by means of biochemical and/or serological testing of the microbial isolates, and then to screen for AMR against a panel of selected antibiotics. Serotyping of the isolates with multiplex polymerase chain reaction revealed the presence of three of the most common clinical Salmonella serovars (S. Enteritidis, S. Typhimurium, S. Thompson) and the major epidemiologically important L. monocytogenes serotypes (1/2a, 1/2b, 1/2c, 4b) in 15 and 9 confirmed isolates of the pathogens, respectively. Strains of Salmonella spp. showed resistance to tetracycline (n = 3) and combined AMR to tetracycline with either ampicillin (n = 2) or trimethoprim-sulfamethoxazole (n = 3), without any multidrug resistance (MDR) being recorded whatsoever. AMR in L. monocytogenes was documented in 55.5% of the bacterial strains (n = 5) tested against ciprofloxacin, meropenem, penicillin, trimethoprim-sulfamethoxazole, and tetracycline. Alarmingly, one strain of L. monocytogenes was MDR to the latter five antibiotics and deemed resistant in three antibiotic groups (carbapenems, penicillins, tetracyclines), after exhibiting minimum inhibitory concentrations (MICs) to meropenem (MIC = 4 μg/mL), penicillin (MIC = 4 μg/mL), and tetracycline (MIC = 48 μg/mL). To the best of our knowledge, finding an MDR L. monocytogenes in pet food is something reported for the first time herein. The results presented in this study highlight the presence of important foodborne bacterial pathogens, such as Salmonella spp. and L. monocytogenes, with increased AMR to antibiotics and possible MDR at the primary production and at the farm level, due to the misuse of pharmacological substances used to treat zoonotic diseases, probably resulting in detection of resistant strains of these pathogenic bacteria in animal-originated food products (e.g., meat, milk, eggs). Full article

This paper reviews recent advances in machine learning (ML) algorithms to improve the postprocessing and interpretation of thermographic data in non-destructive testing (NDT). While traditional NDT methods (e.g., visual inspection, ultrasonic testing) each have their own advantages and limitations, thermographic techniques (e.g., pulsed thermography, laser thermography) have become valuable complementary tools, particularly in inspecting advanced materials such as carbon fiber-reinforced polymers (CFRPs) and superalloys. These techniques generate large volumes of thermal data, which can be challenging to analyze efficiently and accurately. This review focuses on how ML can accelerate defect detection and automated classification in thermographic NDT. We summarize currently popular algorithms and analyze the limitations of existing workflows. Furthermore, this structured analysis provides an in-depth understanding of how artificial intelligence can assist in processing NDT data, with the potential to enable more accurate defect detection and characterization in industrial applications. Full article

This study addresses the nonlinear estimation problem in the strapdown inertial navigation system (SINS) and Doppler velocity log (DVL) integrated navigation by proposing an improved filtering algorithm based on $S{E}_2(3)$ Lie group state representation. A dynamic model satisfying the group affine condition is established to systematically construct both left-invariant and right-invariant error state spaces, upon which two nonlinear filtering approaches are developed. Although the fast unscented transformation method is not novel by itself, its first integration with the $S{E}_2(3)$ Lie group model for SINS/DVL integrated navigation represents a significant advancement. Experimental results demonstrate that under large misalignment angles, the proposed method achieves slightly lower attitude errors compared to linear approaches, while also reducing position estimation errors during dynamic maneuvers. The 12,000 s endurance test confirms the algorithm’s stable long-term performance. Compared with conventional unscented Kalman filter methods, the proposed approach not only reduces computation time by 90% but also achieves real-time processing capability on embedded platforms through optimized sampling strategies and hierarchical state propagation mechanisms. These innovations provide an underwater navigation solution that combines theoretical rigor with engineering practicality, effectively overcoming the computational efficiency and dynamic adaptability limitations of traditional nonlinear filtering methods. Full article

Landslide susceptibility mapping (LSM) faces persistent challenges in defining representative stable samples as conventional random selection often includes unstable areas, introducing spatial bias and compromising model accuracy. To address this, we redefine the certainty factor (CF) method—traditionally for factor weighting—as a spatial screening tool for stable zone delineation and apply it to the tectonically active upper Jinsha River (937 km², southeastern Tibetan Plateau). Our approach first generates a preliminary susceptibility map via CF, using the natural breaks method to define low- and very low-susceptibility zones (CF < 0.1) as statistically stable regions. Non-landslide samples are exclusively selected from these zones for support vector machine (SVM) modeling with five-fold cross-validation. Key results: CF-guided sampling achieves training/testing AUC of 0.924/0.920, surpassing random sampling (0.882/0.878) by 4.8% and reducing ROC standard deviation by 32%. The final map shows 88.49% of known landslides concentrated in 25.70% of high/very high-susceptibility areas, aligning with geological controls (e.g., 92% of high-susceptibility units in soft lithologies within 500 m of faults). Despite using a simpler SVM, our framework outperforms advanced models (ANN: AUC, 0.890; RF: AUC, 0.870) in the same region, proving physical heuristic sample curation supersedes algorithmic complexity. This transferable framework embeds geological prior knowledge into machine learning, offering high-precision risk zoning for disaster mitigation in data-scarce mountainous regions. Full article

Our study tracks the development of 105 Roma children between 3 and 5 (median age: 51 months), enrolled in an NGO-aided developmental program. Each child undergoes pre- and post-assessment based on the Developmental Assessment of Young Children (DAYC), a standard tool used to track the progress in early childhood development and detect delays. Data are gathered from three sources, teacher, parent/caregiver and specialist, covering four developmental domains and adaptive behavior scale. There are subjective biases; however, in the post-assessment, the teachers’ and parents’ evaluations converge. The test results confirm significant improvement in all areas ($p<0.0001$), with the highest being in cognitive skills $32.2\%$ and the lowest being in physical development $14.4\%$. We also apply machine learning methods to impute missing data and predict the likely future progress for a given student in the program based on the initial input, while also evaluating the influence of environmental factors. Our weighted ensemble regression models are coupled with principal component analysis (PCA) and yield average coefficients of determination $R^2\approx 0.7$ for the features of interest. Also, we perform k-means clustering in the plane cognitive vs. social–emotional progress and consider the classification problem of predicting the group in which a given student would eventually be assigned to, with a weighted $F_1$-score of $0.83$ and a macro-averaged area under the curve (AUC) of $0.94$. This could be useful in practice for the optimized formation of study groups. We explore classification as a means of imputing missing categorical data too, e.g., education, employment or marital status of the parents. Our algorithms provide solutions with the $F_1$-score ranging from $0.92$ to $0.97$ and, respectively, an AUC between $0.99$ and 1. Full article

The purpose of the present study was twofold: (i) to examine within-group differences in external load metrics during practice and official competition, and (ii) to examine between-group differences in external load metrics across the U16 and U18 levels of play. A total of thirty-six female athletes participated in the present study, of which nineteen were U16 and seventeen were U18 basketball players. The athletes wore an inertial measurement unit system (Kinexon) sampling at 20 Hz during practice and official games. The average values for each external load metric across ten practices and five games were used for performance analysis. Dependent and independent t-tests were used to examine within- and between-group statistically significant differences, respectively (p < 0.05). The findings reveal that the external load placed on the athletes during the game (e.g., distance covered, average speed, total number of accelerations and decelerations) was considerably greater than the external load during practice sessions, both on the U16 and U18 levels of play. Conversely, while the game-induced external load remained consistent across the two competitive levels, U18 players tended to spend more time and cover more distance in low-speed zones than in high-speed zones during practice, compared to their U16 counterparts, suggesting their superior movement efficiency. Full article

Objective: This study aims to develop and validate an interpretable deep learning framework that leverages self-supervised time reversal (TR) pretraining to identify consistent, biologically plausible functional network biomarkers across multiple neurological and psychiatric disorders. Methods: We pretrained a hierarchical LSTM model using a TR pretext task on the Human Connectome Project (HCP) dataset. The pretrained weights were transferred to downstream classification tasks on five clinical datasets (FBIRN, BSNIP, ADNI, OASIS, and ABIDE) spanning schizophrenia, Alzheimer’s disease, and autism spectrum disorder. After fine-tuning, we extracted latent features and employed a logistic regression probing analysis to decode class-specific functional network contributions. Models trained from scratch without pretraining served as a baseline. Statistical tests (one-sample and two-sample t-tests) were performed on the latent features to assess their discriminative power and consistency. Results: TR pretraining consistently improved classification performance in four out of five datasets, with AUC gains of up to 5.3%, particularly in data-scarce settings. Probing analyses revealed biologically meaningful and consistent patterns: schizophrenia was associated with reduced auditory network activity, Alzheimer’s with disrupted default mode and cerebellar networks, and autism with sensorimotor anomalies. TR-pretrained models produced more statistically significant latent features and demonstrated higher consistency across datasets (e.g., Pearson correlation = 0.9003 for schizophrenia probing vs. −0.67 for non-pretrained). In contrast, non-pretrained models showed unstable performance and inconsistent feature importance. Conclusions: Time Reversal pretraining enhances both the performance and interpretability of deep learning models for fMRI classification. By enabling more stable and biologically plausible representations, TR pretraining supports clinically relevant insights into disorder-specific network disruptions. This study demonstrates the utility of interpretable self-supervised models in neuroimaging, offering a promising step toward transparent and trustworthy AI applications in psychiatry. Full article

As the core component of aircraft systems, aeroengines require accurate Remaining Useful Life (RUL) prediction to ensure flight safety, which serves as a key part of Prognostics and Health Management (PHM). Traditional RUL prediction methods primarily fall into two main categories: physics-based and data-driven approaches. Physics-based methods mainly rely on extensive prior knowledge, limiting their scalability, while data-driven methods (including statistical analysis and machine learning) struggle with handling high-dimensional data and suboptimal modeling of multi-scale temporal dependencies. To address these challenges and enhance prediction accuracy and robustness, we propose a novel hybrid deep learning framework (CLSTM-TCN) integrating 2D Convolutional Neural Network (2D-CNN), Long Short-Term Memory (LSTM) network, and Temporal Convolutional Network (TCN) modules. The CLSTM-TCN framework follows a progressive feature refinement logic: 2D-CNN first extracts short-term local features and inter-feature interactions from input data; the LSTM network then models long-term temporal dependencies in time series to strengthen global temporal dynamics representation; and TCN ultimately captures multi-scale temporal features via dilated convolutions, overcoming the limitations of the LSTM network in long-range dependency modeling while enabling parallel computing. Validated on the NASA C-MAPSS data set (focusing on FD001), the CLSTM-TCN model achieves a root mean square error (RMSE) of 13.35 and a score function (score) of 219. Compared to the CNN-LSTM, CNN-TCN, and LSTM-TCN models, it reduces the RMSE by 27.94%, 30.79%, and 30.88%, respectively, and significantly outperforms the traditional single-model methods (e.g., standalone CNN or LSTM network). Notably, the model maintains stability across diverse operational conditions, with RMSE fluctuations capped within 15% for all test cases. Ablation studies confirm the synergistic effect of each module: removing 2D-CNN, LSTM, or TCN leads to an increase in the RMSE and score. This framework effectively handles high-dimensional data and multi-scale temporal dependencies, providing an accurate and robust solution for aeroengine RUL prediction. While current performance is validated under single operating conditions, ongoing efforts to optimize hyperparameter tuning, enhance adaptability to complex operating scenarios, and integrate uncertainty analysis will further strengthen its practical value in aircraft health management. Full article