Search Results (5,283)

Assistance dogs have shown promise in alerting to epileptic seizures in their owners, but current approaches often lack consistency, standardisation, and objective validation. This proof-of-concept study presents the development and initial validation of a wearable behaviour-alert detection collar developed for trained assistance dogs. It demonstrates the technical feasibility for automated detection of trained signalling behaviours. The collar integrates an inertial sensor and machine learning pipeline to detect a specific, trained alert behaviour of two rapid clockwise spins used by dogs to signal a seizure event. Data were collected from six trained dogs, resulting in 135 labelled spin alerts. Although the dataset size is limited compared to other machine learning applications, this reflects the real-world constraint that it is not practical for assistance dogs to perform excessive spin signalling during their training. Four supervised machine learning models (Random Forest, Logistic Regression, Naïve Bayes, and SVM) were evaluated on segmented accelerometer and gyroscope data. Random Forest achieved the highest performance (F1-score = 0.65; accuracy = 92%) under a Leave-One-DOG-Out (LODO) protocol. The system represents a novel step toward combining intentional canine behaviours with wearable technology, aligning with trends on the Internet of Medical Things. This proof-of-concept demonstrates technical feasibility and provides a foundation for future development of real-time seizure-alerting systems, representing an important first step toward scalable animal-assisted healthcare innovation. Full article

Despite nationwide control efforts, central China experiences persistently high annual PM_2.5 concentrations (~50 μg/m³), which are particularly severe in January (exceeding 110 μg/m³). This study employs an integrated approach combining a Multiple Linear Regression (MLR) model derived from random forest analysis with the WRF-CMAQ chemical transport modeling system to quantitatively disentangle the driving factors of PM_2.5 concentrations in central China. Key findings reveal significant spatiotemporal heterogeneity in anthropogenic contributions, evidenced by consistently higher north–south gradients in regression residuals (reflecting emission impacts), linked to spatially varying industrial and transportation influences. Critically, the reduction in anthropogenic impacts over six years was substantially smaller in winter (January: 27 to 23 μg/m³) compared to summer (15 to −18 μg/m³, July), highlighting the profound role of emissions in driving severe January pollution events. Furthermore, WRF-CMAQ simulations demonstrated that adverse meteorological conditions in January 2020 counteracted emission controls, causing a net increase in PM_2.5 of +13 μg/m³ relative to 2016, thereby offsetting ~68% of the reductions achieved through emission abatement (−19 μg/m³). Significant regional transport, especially affecting northern and central Henan, further weakened local control efficacy. These quantitative insights into the mechanisms of PM_2.5 pollution, particularly the counteracting effects of meteorology on emission reductions in critical winter periods, provide a vital scientific foundation for designing more effective and targeted air quality management strategies in central China. Full article

The increasing demand for monoclonal antibody (mAb) therapeutics has intensified the need for more efficient and consistent biomanufacturing processes. We present a data-driven, machine-learning (ML) approach to exploring and predicting upstream yield behavior. Drawing on industrial-scale batch records for a single mAb product from a contract development and manufacturing organization, we applied regression models to identify key process parameters and estimate production outcomes. Random forest regression, gradient boosting machine, and support vector regression (SVR) were evaluated to predict three yield indicators: bioreactor final weight (BFW), harvest titer (HT), and packed cell volume (PCV). SVR outperformed other models for BFW prediction (R² = 0.978), while HT and PCV were difficult to model accurately with the available data. Exploratory analysis using sequential least-squares programming suggested parameter combinations associated with improved yield estimates relative to historical data. Sensitivity analysis highlighted the most influential process parameters. While the findings demonstrate the potential of ML for predictive, data-driven yield improvement, the results should be interpreted as an exploratory proof of concept rather than a fully validated optimization framework. This study highlights the need to incorporate process constraints and control logic, along with interpretable or hybrid modeling frameworks, to enable practical deployment in regulated biomanufacturing environments. Full article

This paper investigates the short-term predictability of daily crude oil price movements by employing a multi-method analytical framework that incorporates both econometric and machine learning techniques. Utilizing a dataset of 21 financial and commodity time series spanning ten years of trading days (2015–2024), we explore the dynamics of oil price volatility and its key determinants. In the forecasting phase, we applied seven models. The meta-learner model, which consists of three base learners (Random Forest, gradient boosting, and support vector regression), achieved the highest R² value of 0.532, providing evidence that our complex model structure can successfully outperform existing approaches. This ensemble demonstrated that the most influential predictors of next-day oil prices are VIX, OVX, and MOVE (volatility indices for equities, oil, and bonds, respectively), and lagged oil returns. The results underscore the critical role of volatility spillovers and nonlinear dependencies in forecasting oil returns and suggest future directions for integrating macroeconomic signals and advanced volatility models. Moreover, we show that combining multiple machine learning procedures into a single meta-model yields superior predictive performance. Full article

High-resolution air temperature (AT) data is essential for understanding urban heat dynamics, particularly in urban areas characterised by complex microclimates. However, AT is rarely available in such detail, emphasising the need for its modelling. This study employs a Random Forest regression framework to predict 20 m resolution AT maps across Milan, Italy, for 2022. We focus on seasonal heatwave periods, identified from a long-term climate reanalysis dataset, and multiple diurnal and nocturnal phases that reflect the daily evolution of AT. We predict AT from a high-quality dataset of 97 authoritative and crowdsourced stations, incorporating predictors derived from geospatial and Earth Observation data, including Sentinel-2 indices and urban morphology metrics. Model performance is highest during the late afternoon and nighttime, with an average $R^2$ between 0.33 and 0.37, and an RMSE between 0.7 and 1.4 °C. This indicates modest, yet reasonable agreement with observations, given the challenges of high-resolution AT mapping. Daytime predictions prove more challenging, as noted in previous studies using similar methods. Furthermore, we explore the potential of hyperspectral (HS) data to estimate surface material abundances through spectral unmixing and assess their influence on AT. Results highlight the added value of HS-derived material abundance maps for insights into urban thermal properties and their relationship with AT patterns. The produced maps are useful for identifying intra-urban AT variability during extreme heat conditions and can support numerical model validation and city-scale heat mitigation planning. Full article

In recent years, ecosystems in Saudi Arabia have experienced severe degradation due to factors such as hyperaridity, overgrazing, climate change, urban expansion, and an increase in uncontrolled wildfires. Among these, wildfires have emerged as the second most significant threat to forests after urban expansion. This study aims to map wildfire susceptibility in southwestern Saudi Arabia by identifying key driving factors and evaluating the performance of several machine learning models under conditions of limited and imbalanced data. The models tested include Maxent, logistic regression, random forest, XGBoost, and support vector machine. In addition, an NDVI-based phenological approach was applied to assess seasonal vegetation dynamics and to compare its effectiveness with conventional machine learning-based susceptibility mapping. All methods generated effective wildfire risk maps, with Maxent achieving the highest predictive accuracy (AUC = 0.974). The results indicate that human activities and dense vegetation cover are the primary contributors to wildfire occurrence. This research provides valuable insights for wildfire risk assessment in data-scarce regions and supports proactive fire management strategies in non-traditional fire-prone environments. Full article

Postpartum haemorrhage is one of the leading causes of maternal morbidity and mortality worldwide. The early identification of bleeding risk in individual women is crucial for enabling timely interventions and improving patient outcomes.This study aims to evaluate various exploratory and classification methodologies, alongside optimisation strategies, for identifying predictors of postpartum haemorrhage. K-means clustering was employed on a retrospective cohort of patients, incorporating demographic, obstetric, and laboratory variables, to delineate patient profiles and select pertinent features. Initially, a classical logistic regression model, implemented without cross-validation, facilitated the identification of six significant predictors for postpartum haemorrhage: lactate dehydrogenase, urea, platelet count, non-O blood group, gestational age, and first-degree lacerations, all of which are variables routinely available in clinical practice. Furthermore, machine learning algorithms—including stepwise logistic regression, ridge logistic regression, and random forest—were utilised, applying cross-validation to optimise predictive performance and enhance generalisability. Among these methodologies, ridge logistic regression emerged as the most effective model, achieving the following metrics: sensitivity 0.857, specificity 0.875, accuracy 0.871, F1-score 0.759, and AUC 0.907. While machine learning techniques demonstrated superior performance, the integration of classical statistical methods with machine learning approaches provides a robust framework for generating reliable predictions and fostering significant clinical insights. Full article

Background: Comprehensive data on the economic burden of cervical cancer treatment remain scarce in China’s less developed regions, necessitating this study on hospitalization costs and expenditure trends in these areas. Methods: Employing a multi-stage stratified cluster sampling approach, this study enrolled 10,070 cervical cancer inpatients from 72 healthcare facilities in Gansu Province. Clinical and expenditure data were extracted from hospital information systems. Rank sum tests and Spearman correlation analyses were performed for univariate assessment, while quantile regression and random forest models were applied to identify determinant factors. Results: From 2019 to 2023, the average hospitalization duration for cervical cancer patients in Gansu Province was 16.12 days, with an average hospitalization cost of USD 3862.08 (2023 constant prices, converted from CNY at 1:7.0467). During these five years, the average inpatient costs per hospitalization increased from USD 3473.45 to USD 4202.57, and the average daily hospitalization cost rose from USD 230.53 to USD 241.77. The average drug cost decreased from USD 769.06 to USD 640.16. The main factors influencing hospitalization costs included the length of hospital stay, whether cervical cancer surgery was performed, hospital type, hospital level, and the proportion of medications. Conclusions: Our findings indicate that cervical cancer is a considerable economic burden on both families and society. This highlights the need to control the length of hospital stay and optimize the allocation of medical resources, in addition to strengthening cervical cancer screening and HPV vaccination in underdeveloped areas, in order to enhance the efficiency of prevention and treatment and ensure medical equity. Full article

Background: The automated analysis of intestinal organoids in microscopy images are essential for high-throughput morphological studies, enabling precision and scalability. Traditional manual analysis is time-consuming and subject to observer bias, whereas Machine Learning (ML) approaches have recently demonstrated superior performance. Purpose: This study aims to evaluate YOLO (You Only Look Once) for organoid segmentation and classification, comparing its standalone performance with a hybrid pipeline that integrates DL-based feature extraction and ML classifiers. Methods: The dataset, consisting of 840 light microscopy images and over 23,000 annotated intestinal organoids, was divided into training (756 images) and validation (84 images) sets. Organoids were categorized into four morphological classes: cystic non-budding organoids (Org0), early organoids (Org1), late organoids (Org3), and Spheroids (Sph). YOLO version 10 (YOLOv10) was trained as a segmenter-classifier for the detection and classification of organoids. Performance metrics for YOLOv10 as a standalone model included Average Precision (AP), mean AP at 50% overlap (mAP50), and confusion matrix evaluated on the validation set. In the hybrid pipeline, trained YOLOv10 segmented bounding boxes, and features extracted from these regions using YOLOv10 and ResNet50 were classified with ML algorithms, including Logistic Regression, Naive Bayes, K-Nearest Neighbors (KNN), Random Forest, eXtreme Gradient Boosting (XGBoost), and Multi-Layer Perceptrons (MLP). The performance of these classifiers was assessed using the Receiver Operating Characteristic (ROC) curve and its corresponding Area Under the Curve (AUC), precision, F1 score, and confusion matrix metrics. Principal Component Analysis (PCA) was applied to reduce feature dimensionality while retaining 95% of cumulative variance. To optimize the classification results, an ensemble approach based on AUC-weighted probability fusion was implemented to combine predictions across classifiers. Results: YOLOv10 as a standalone model achieved an overall mAP50 of 0.845, with high AP across all four classes (range 0.797–0.901). In the hybrid pipeline, features extracted with ResNet50 outperformed those extracted with YOLO, with multiple classifiers achieving AUC scores ranging from 0.71 to 0.98 on the validation set. Among all classifiers, Logistic Regression emerged as the best-performing model, achieving the highest AUC scores across multiple classes (range 0.93–0.98). Feature selection using PCA did not improve classification performance. The AUC-weighted ensemble method further enhanced performance, leveraging the strengths of multiple classifiers to optimize prediction, as demonstrated by improved ROC-AUC scores across all organoid classes (range 0.92–0.98). Conclusions: This study demonstrates the effectiveness of YOLOv10 as a standalone model and the robustness of hybrid pipelines combining ResNet50 feature extraction and ML classifiers. Logistic Regression emerged as the best-performing classifier, achieving the highest ROC-AUC across multiple classes. This approach ensures reproducible, automated, and precise morphological analysis, with significant potential for high-throughput organoid studies and live imaging applications. Full article

Cultivated land (CL) is essential for human survival, as its coordinated utilization plays a crucial role in both food production and ecological protection. In this study, we focus on Aksu, a typical oasis in arid areas of Xinjiang, to explore how to improve the eco-efficiency of cultivated land utilization (ECLU) from the perspective of carbon emissions under different ownership structures. The goal is to provide policy support for the sustainable intensification of CL in Aksu. The super-efficiency slack-based measure (Super-SBM) model was used to calculate the ECLU, while the carbon emissions coefficient method was employed to estimate cultivated land carbon emissions (CLCE). Additionally, the random forest regression (RFR) model was utilized to analyze differences in CLCE between collective and state-owned cultivated lands. Finally, a Geo-detector analysis was conducted to identify driving factors of CLCE. The findings indicate that the overall ECLU values in Aksu initially increased and subsequently decreased over time. During the study period, Kalpin showed the highest ECLU, followed by Wensu and Wushi. The total CLCE in Aksu demonstrated an initial increase followed by a decrease, but the overall trend was growth, from 3.7 t in 2008 to 5.63 t in 2019, on average. It was observed that carbon emissions from state-owned cultivated land were greater than those from collective cultivated land, and carbon emissions from non-food crops were higher than those from food crops. Furthermore, spatial heterogeneity was evident in the CLCE. The single factor detection results showed that the Local_GDP (q = 0.763, representing the explanatory power of the Local_GDP on cultivated land carbon emissions) was identified as the main driver of CLCE in Aksu. The interactive detection results indicated that the Local_GDP and Farmer income (0.839) had stronger effects on CLCE in Aksu than any other two factors. It was also found that ownership of CL directly affects CLCE and indirectly affects the ECLU. In conclusion, it is necessary to formulate corresponding countermeasures for improving the ECLU involving government intervention, as well as cooperation with farmers and other stakeholders, to address these issues effectively within Aksu’s agricultural sector. Full article

Compaction quality control in earthworks and pavements still relies mainly on density-based acceptance referenced to laboratory Proctor tests, which are costly, time-consuming, and spatially sparse. Lightweight dynamic cone penetrometer (LDCP) provides rapid indices, such as $q_{d0}$ and $q_{d1}$, yet acceptance thresholds commonly depend on ad hoc, site-specific calibrations. This study develops and validates a supervised machine learning framework that estimates $q_{d0}$, $q_{d1}$, and $Z_c$ directly from readily available soil descriptors (gradation, plasticity/activity, moisture/state variables, and GTR class) using a multi-campaign dataset of $n=360$ observations. While the framework does not remove the need for the standard soil characterization performed during design (e.g., W, ${\gamma }_{d,\mathrm{field}}$, and $R{C}_{\mathrm{SPC}}$), it reduces reliance on additional LDCP calibration campaigns to obtain device-specific reference curves. Models compared under a unified pipeline include regularized linear baselines, support vector regression, Random Forest, XGBoost, and a compact multilayer perceptron (MLP). The evaluation used a fixed 80/20 train–test split with 5-fold cross-validation on the training set and multiple error metrics ($R^2$, RMSE, MAE, and MAPE). Interpretability combined SHAP with permutation importance, 1D partial dependence (PDP), and accumulated local effects (ALE); calibration diagnostics and split-conformal prediction intervals connected the predictions to QA/QC decisions. A naïve GTR-average baseline was added for reference. Computation was lightweight. On the test set, the MLP attained the best accuracy for $q_{d1}$ ($R^2=0.794$, RMSE $=5.866$), with XGBoost close behind ($R^2=0.773$, RMSE $=6.155$). Paired bootstrap contrasts with Holm correction indicated that the MLP–XGBoost difference was not statistically significant. Explanations consistently highlighted density- and moisture-related variables (${\gamma }_{d,\mathrm{field}}$, $R{C}_{\mathrm{SPC}}$, and W) as dominant, with gradation/plasticity contributing second-order adjustments; these attributions are model-based and associational rather than causal. The results support interpretable, computationally efficient surrogates of LDCP indices that can complement density-based acceptance and enable risk-aware QA/QC via conformal prediction intervals. Full article

Rapid decarbonization relies on knowing which structural and energy factors affect national carbon dioxide emissions. Much of the literature leans on linear and additive assumptions, which may gloss over curvature and interactions in this energy–emissions link. Unlike previous studies, we take a different approach. Using a panel of 80 high- and upper-middle-income countries from 2011 to 2020, we model emissions as a function of fossil fuel energy consumption, methane, the food production index, renewable electricity output, gross domestic product (GDP), and trade measured as trade over GDP. Our contribution is twofold. First, we evaluate how different modeling strategies, from a traditional Generalized Linear Model to more flexible approaches such as Support Vector Machine regression and Random Forest (RF), influence the identification of emission drivers. Second, we use Double Machine Learning (DML) to estimate the incremental effect of fossil fuel consumption while controlling for other variables, offering a more careful interpretation of its likely causal role. Across models, a clear pattern emerges: GDP dominates; fossil fuel energy consumption and methane follow. Renewable electricity output and trade contribute, but to a moderate degree. The food production index adds little in this aggregate, cross-country setting. To probe the mechanism rather than the prediction, we estimate the incremental role of fossil fuel energy consumption using DML with RF nuisance functions. The partial effect remains positive after conditioning on the other covariates. Taken together, the results suggest that economic scale and the fuel mix are the primary levers for near-term emissions profiles, while renewables and trade matter, just less than is often assumed and in ways that may depend on context. Full article

The proliferation of Internet of Medical Things (IoMT) devices connected to the internet poses significant challenges to data integrity, confidentiality, and patient safety due to their vulnerability to outside exploitation. Specifically, IoMT devices capture and process vast amounts of sensitive patient data but often lack adequate security mechanisms, making them susceptible to attacks that compromise data integrity—such as the injection of false or fabricated information—which imposes significant risks on the patient. To address this, we introduce a novel hybrid anomaly detection model combining a Graph Convolutional Network (GCN) with a transformer architecture. The GCN captures the structural relationships within the IoMT data, while the transformer models the sequential dependencies in the anomalies. We evaluate our approach using the novel CICIOMT24 dataset, the first of its kind to emulate real-world IoMT network traffic from over 40 devices and 18 distinct cyberattacks. Compared against several machine learning baselines (including Logistic Regress, Random Forest, and Adaptive Boosting), the hybrid model effectively captures attacks and provides early detection capabilities. This work demonstrates a scalable and robust solution to enhance the safety and security of both IoMT devices and critical patient data. Full article

Cost is a key performance indicator for evaluating the success of architectural construction projects. While previous studies have relied on quantitative data and statistical models to evaluate cost performance, recent advancements in methods have enabled analysis using unstructured data. Unstructured data, particularly in construction supervision reports, can be considered the significant variables for performance evaluation, as they represent independent third-party monitoring of the construction project’s execution. This study aims to present a framework that supports cost performance evaluation using unstructured data and random forests (RFs), a representative method of machine learning. Specifically, association rule analysis and social network analysis were used to identify the main keywords, and an RF model was applied to these data to evaluate cost performance. The tuning of hyper-parameters in the RF was implemented by the Bayesian optimization technique with the augmentation of the original dataset. The accuracy of cost performance evaluation was 59% for the traditional logistic regression (LR), 74% for the regularization-based logistic regression (BLR) designed to prevent overfitting, and 76% for the RF model utilizing augmented data. The complementary utility of the models consisting of the proposed framework can be useful for deriving various evaluation explanations about cost performance. The applicability is expected to increase as more data become available in the future. Full article

The construction industry is a major contributor to global resource consumption and waste. This sector extracts over two billion tons of raw materials each year and contributes over 30% of all solid waste generated annually through construction and demolition debris. The movement toward circularity in the built environment aims to replace linear processes of extraction and disposal by promoting policies favoring building preservation and adaptive reuse, as well as the salvage and reuse of building materials. Few North American cities have implemented explicit policies that incentivize circularity to decouple urban growth from resource consumption, and there remain substantial hurdles to adoption. Nonetheless, existing regulatory and planning tools, such as zoning codes and historic preservation policies, may already influence redevelopment in ways that could align with circularity. This article examines spatial patterns in these indirect pathways through a case study of a college town in New York State, assessing how commonly used local planning tools shape urban redevelopment trajectories. Using a three-stage spatial analysis protocol, including exploratory analysis, Geographically Weighted Regressions (GWRs), and Geographic Random Forest (GRF) modeling, the study evaluates the impact of zoning regulations and historic preservation designations on patterns of demolition, reinvestment, and incremental change in the building stock. National historic districts were strongly associated with more building adaptation permits indicating reinvestment in existing buildings. Mixed-use zoning was positively correlated with new construction, while special overlay districts and low-density zoning were mostly negatively correlated with concentrations of building adaptation permits. A key contribution of this paper is a replicable protocol for urban building stock analysis and insights into how land use policies can support or hinder incremental urban change in moves toward the circular city. Further, we provide recommendations for data management strategies in small cities that could help strengthen analysis-driven policies. Full article