Search Results (2,797)

The accurate forecasting of surface air temperature (T2M) is crucial for climate analysis, agricultural planning, and energy management. This study proposes a novel forecasting framework grounded in structured temporal decomposition. Using the Kolmogorov–Zurbenko (KZ) filter, all predictor variables are decomposed into three physically interpretable components: long-term, seasonal, and short-term variations, forming an expanded multi-scale feature space. A central innovation of this framework lies in training a single unified model on the decomposed feature set to predict the original target variable, thereby enabling the direct learning of scale-specific driver–response relationships. We present the first comprehensive benchmarking of this architecture, demonstrating that it consistently enhances the performance of both regularized linear models (Ridge and Lasso) and tree-based ensemble methods (Random Forest and XGBoost). Under rigorous walk-forward validation, the framework substantially outperforms conventional, non-decomposed approaches—for example, XGBoost improves the coefficient of determination ($R^2$) from 0.80 to 0.91. Furthermore, temporal decomposition enhances interpretability by enabling Ridge and Lasso models to achieve performance levels comparable to complex ensembles. Despite these promising results, we acknowledge several limitations: the analysis is restricted to a single geographic location and time span, and short-term components remain challenging to predict due to their stochastic nature and the weaker relevance of predictors. Additionally, the framework’s effectiveness may depend on the optimal selection of KZ parameters and the availability of sufficiently long historical datasets for stable walk-forward validation. Future research could extend this approach to multiple geographic regions, longer time series, adaptive KZ tuning, and specialized short-term modeling strategies. Overall, the proposed framework demonstrates that temporal decomposition of predictors offers a powerful inductive bias, establishing a robust and interpretable paradigm for surface air temperature forecasting. Full article

Reference evapotranspiration (ET₀) is a fundamental variable for irrigation scheduling and water management. Conventional estimation methods, such as the FAO-56 Penman–Monteith equation, are of limited use in developing regions where meteorological data are scarce. This study evaluates the potential of machine learning (ML) approaches to estimate ET₀ in Paraguay, using only geographical and temporal predictors—latitude, longitude, altitude, and month. Five algorithms were tested: artificial neural networks (ANNs), k-nearest neighbors (KNN), random forest (RF), extreme gradient boosting (XGB), and adaptive neuro-fuzzy inference systems (ANFISs). The framework consisted of ET₀ calculation, baseline model testing (ML techniques), ensemble modeling, leave-one-station-out validation, and spatial interpolation by inverse distance weighting. ANFIS achieved the highest prediction accuracy (R² = 0.950, RMSE = 0.289 mm day⁻¹, MAE = 0.202 mm day⁻¹), while RF and XGB showed stable and reliable performance across all stations. Spatial maps highlighted strong seasonal variability, with higher ET₀ values in the Chaco region in summer and lower values in winter. These results confirm that ML algorithms can generate robust ET₀ estimates under data-constrained conditions, and provide scalable and cost-effective solutions for irrigation management and agricultural planning in Paraguay. Full article

Accurate segmentation of crops and weeds is essential for enhancing crop yield, optimizing herbicide usage, and mitigating environmental impacts. Traditional weed management practices, such as manual weeding or broad-spectrum herbicide application, are labor-intensive, environmentally harmful, and economically inefficient. In response, this study introduces a novel precision agriculture framework integrating Unmanned Aerial Vehicle (UAV)-based remote sensing with advanced deep learning techniques, combining Super-Resolution Reconstruction (SRR) and semantic segmentation. This study is the first to integrate UAV-based SRR and semantic segmentation for tobacco fields, systematically evaluate recent Transformer and Mamba-based models alongside traditional CNNs, and release an annotated dataset that not only ensures reproducibility but also provides a resource for the research community to develop and benchmark future models. Initially, SRR enhanced the resolution of low-quality UAV imagery, significantly improving detailed feature extraction. Subsequently, to identify the optimal segmentation model for the proposed framework, semantic segmentation models incorporating CNN, Transformer, and Mamba architectures were used to differentiate crops from weeds. Among evaluated SRR methods, RCAN achieved the optimal reconstruction performance, reaching a Peak Signal-to-Noise Ratio (PSNR) of 24.98 dB and a Structural Similarity Index (SSIM) of 69.48%. In semantic segmentation, the ensemble model integrating Transformer (DPT with DINOv2) and Mamba-based architectures achieved the highest mean Intersection over Union (mIoU) of 90.75%, demonstrating superior robustness across diverse field conditions. Additionally, comprehensive experiments quantified the impact of magnification factors, Gaussian blur, and Gaussian noise, identifying an optimal magnification factor of 4×, proving that the method was robust to common environmental disturbances at optimal parameters. Overall, this research established an efficient, precise framework for crop cultivation management, offering valuable insights for precision agriculture and sustainable farming practices. Full article

Surrogate-assisted evolutionary algorithms (SAEAs) have emerged as a powerful class of optimization methods that utilize surrogate models to address expensive optimization problems (EOPs), where fitness evaluations (FEs) are expensive or limited. By leveraging previously evaluated solutions to learn predictive models, SAEAs enable efficient search under constrained evaluation budgets. However, the performance of SAEAs heavily depends on the quality and utilization of surrogate models, and balancing the accuracy and generalization ability makes effective model construction and management a key challenge. Therefore, this paper introduces a novel probability selection-based surrogate-assisted evolutionary algorithm (PS-SAEA) to enhance optimization performance under FE-constrained conditions. The PS-SAEA has two novel designs. First, a probabilistic model selection (PMS) strategy is proposed to stochastically select surrogate models, striking a balance between prediction accuracy and generalization by avoiding overfitting commonly caused by greedy selection. Second, a weighted model ensemble (WME) mechanism is developed to integrate selected models, assigning weights based on individual prediction errors to improve the accuracy and reliability of fitness estimation. Extensive experiments on benchmark problems with varying dimensionalities demonstrate that PS-SAEA consistently outperforms several state-of-the-art SAEAs, validating its effectiveness and robustness in dealing with various complex EOPs. Full article

Water loss during storage is a major cause of postharvest quality deterioration in cucumber, yet existing methods to monitor hydration are often destructive or require expensive instrumentation. We developed a low-cost, non-destructive approach for estimating fruit relative water content (RWC) using visible-light color imaging combined with an ensemble machine-learning model (Random Forest). A total of 1200 fruits were greenhouse-grown, harvested at market maturity, and equally divided between optimal and ambient storage temperature (10 and 25 °C, respectively). Digital images were acquired at harvest and at 7 d intervals during storage, and color parameters from four standard color systems (RGB, CMYK, CIELAB, HSV) were extracted separately for the neck, mid, and blossom regions as well as for the whole fruit. During storage, fruit RWC decreased from 100% (fully hydrated condition) to 15.3%, providing a broad dynamic range for assessing color–hydration relationships. Among the 16 color features evaluated, the mean cyan component (μC) of the CMYK space showed the strongest relationship with measured RWC (R² up to 0.70 for whole-fruit averages), reflecting the cyan region’s heightened sensitivity to dehydration-induced changes in pigments, cuticle properties and surface scattering. The Random Forest regression model trained on these features achieved a higher predictive accuracy (R² = 0.89). Predictive accuracy was also consistently higher when μC was calculated over the entire fruit surface rather than for individual anatomical regions, indicating that whole-fruit color information provides a more robust hydration signal than region-specific measurements. Our findings demonstrate that simple visible-range imaging coupled with ensemble learning can provide a cost-effective, non-invasive tool for monitoring postharvest hydration of cucumber fruit, with direct applications in quality control, shelf-life prediction and waste reduction across the fresh-produce supply chain. Full article

Background/Objectives: This study presents an artificial intelligence (AI)-supported incentive spirometry system designed to explore the feasibility of automating the monitoring of respiratory exercises, a critical nursing intervention for maintaining pulmonary function and reducing postoperative complications. Methods: This system uses a tablet’s camera to track a standard spirometer’s volume indicator in real-time, reducing the manual nursing workload, unlike traditional mechanical spirometers that lack feedback capabilities. Image processing techniques analyze exercise performance, while the interface provides instant feedback, data recording, and graphical display. Machine learning models (Random Forest, XGBoost, Gradient Boosting, SVM, Logistic Regression, KNN) were trained on scripted patient data, including demographics, smoking status, and spirometry measurements, to classify respiratory performance as “poor”, “good”, or “excellent”. Results: The ensemble methods demonstrated exceptional performance, achieving 100% accuracy and R² = 1.0, with cross-validation mean accuracies exceeding 99.4%. This feasibility study demonstrates the technical viability of this AI-driven approach and lays the groundwork for future clinical validation. Conclusions: This system presents a potential cost-effective, accessible solution suitable for both clinical and home settings, potentially integrating into standard respiratory care protocols. This system not only reduces nursing workload but also has the potential to improve patient adherence. This pilot study demonstrates the technical feasibility and potential of this AI-driven approach, laying the groundwork for future clinical validation. Full article

In recent years, academic performance prediction has evolved as a research field thanks to its development and exploration in the educational context. Early student performance prediction is crucial for enhancing educational outcomes and implementing timely interventions. Conventional approaches frequently struggle on behalf of the complexity of student profiles as a consequence of single activation functions, which prevent them from effectively learning intricate patterns. In addition, these models could experience obstacles such as the vanishing gradient problem and computational complexity. Therefore, this research study designed an Activation Ensemble Deep Neural Network (AcEnDNN) model to gain control of the previously mentioned challenges. The main contribution is the creation of a credible student performance prediction model that comprises extensive data preprocessing, feature extraction, and an Activation Ensemble DNN. By utilizing various methods of activation functions, such as ReLU, tanh, sigmoid, and swish, the ensembled activation functions are able to learn the complex structure of student data, which leads to more accurate performance prediction. The AcEn-DNN model is trained and evaluated based on the publicly available Student-mat.csv dataset, Student-por.csv dataset, and a real-time dataset. The experimental results revealed that the AcEn-DNN model achieved lower error rates, with an MAE of 1.28, MAPE of 2.36, MSE of 4.55, and RMSE of 2.13 based on a training percentage of 90%, confirming its robustness in modeling nonlinear relationships within student data. The proposed model also gained the minimum error values MAE of 1.28, MAPE of 2.97, MSE of 4.77, and RMSE of 2.18, based on a K-fold value of 10, utilizing the Student-mat.csv dataset. These findings highlight the model’s potential in early identification of at-risk students, enabling educators to develop targeted learning strategies. This research contributes to educational data mining by advancing predictive modeling techniques that evaluate student performance. Full article

Background/Objectives: Lung cancer remains the leading cause of cancer-related mortality worldwide among both men and women. Early and accurate detection is essential to improve patient outcomes. This study explores the use of artificial intelligence (AI)-based software for the diagnosis of lung cancer through the analysis of medical images in DICOM format, aiming to enhance image visualization, preprocessing, and diagnostic precision in chest computed tomography (CT) scans. Methods: The proposed system processes DICOM medical images converted to standard formats (JPG or PNG) for preprocessing and analysis. An ensemble of classical machine learning algorithms—including Random Forest, Gradient Boosting, Support Vector Machine, and K-Nearest Neighbors—was implemented to classify pulmonary images and predict the likelihood of malignancy. Image normalization, denoising, segmentation, and feature extraction were performed to improve model reliability and reproducibility. Results: The AI-enhanced system demonstrated substantial improvements in diagnostic accuracy and robustness compared with individual classifiers. The ensemble model achieved a classification accuracy exceeding 90%, highlighting its effectiveness in identifying malignant and non-malignant lung nodules. Conclusions: The findings indicate that AI-assisted CT image processing can significantly contribute to the early detection of lung cancer. The proposed methodology enhances diagnostic confidence, supports clinical decision-making, and represents a viable step toward integrating AI into radiological workflows for early cancer screening. Full article

Objectives: This paper introduces an adaptive learning framework for handling concept drift in data by dynamically adjusting model updates based on the severity of detected drift. Methods: The proposed method combines multiple statistical measures to quantify distributional changes between recent and historical data windows. The resulting severity score drives a three-tier adaptation policy: minor drift is ignored, moderate drift triggers incremental model updates, and severe drift initiates full model retraining. Results: This approach balances stability and adaptability, reducing unnecessary computation while preserving model accuracy. The framework is applicable to both single-model and ensemble-based systems, offering a flexible and efficient solution for real-time drift management. Also, different transformation methods were reviewed, and quantile transformation was tested. By applying a quantile transformation, the Kolmogorov–Smirnov (KS) statistic decreased from 0.0559 to 0.0072, demonstrating effective drift adaptation. 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

As modern power systems integrate more renewable and decentralised generation, maintaining grid stability has become increasingly challenging. This study proposes a data-driven machine learning framework for forecasting system inertia service costs—a key yet underexplored variable influencing energy trading and frequency stability in Great Britain. Using eight years (2017–2024) of National Energy System Operator (NESO) data, four models—Long Short-Term Memory (LSTM), Residual LSTM, eXtreme Gradient Boosting (XGBoost), and Light Gradient-Boosting Machine (LightGBM)—are comparatively analysed. LSTM-based models capture temporal dependencies, while ensemble methods effectively handle nonlinear feature relationships. Results demonstrate that LightGBM achieves the highest predictive accuracy, offering a robust method for inertia cost estimation and market intelligence. The framework contributes to strategic procurement planning and supports market design for a more resilient, cost-effective grid. Full article

Land use classification based on multi-modal data fusion has gained significant attention due to its potential to capture the complex characteristics of urban environments. However, effectively extracting and integrating discriminative features derived from heterogeneous geospatial data remain challenging. This study proposes an ensemble learning framework for land use classification by fusing cross-modal deep features from both physical and socioeconomic perspectives. Specifically, the framework utilizes the Masked Autoencoder (MAE) to extract global spatial dependencies from remote sensing imagery and applies long short-term memory (LSTM) networks to model spatial distribution patterns of points of interest (POIs) based on type co-occurrence. Furthermore, we employ inter-modal contrastive learning to enhance the representation of physical and socioeconomic features. To verify the superiority of the ensemble learning framework, we apply it to map the land use distribution of Bejing. By coupling various physical and socioeconomic features, the framework achieves an average accuracy of 84.33 %, surpassing several comparative baseline methods. Furthermore, the framework demonstrates comparable performance when applied to a Shenzhen dataset, confirming its robustness and generalizability. The findings highlight the importance of fully extracting and effectively integrating multi-source deep features in land use classification, providing a robust solution for urban planning and sustainable development. Full article

Assessing and forecasting hydrocarbon resource potential (HRP) is of great significance. However, due to the complexity and uncertainty of geological conditions during hydrocarbon accumulation, it is challenging to accurately establish HRP models. This study employs machine learning methods to construct a HRP assessment model. First, nine primary controlling factors were selected from the five key conditions for HRP: source rock, reservoir, trap, migration, and accumulation. Subsequently, three prediction models were developed based on the backpropagation (BP) neural network, BP-Bagging algorithm, and BP-AdaBoost algorithm, with hydrocarbon resources abundance as the output metric. These models were applied to the Dongpu Depression in the Bohai Bay Basin for performance evaluation and optimization. Finally, this study examined the importance of various variables in predicting HRP and analyzed model uncertainty. The results indicate that the BP-AdaBoost model outperforms the others. On the test dataset, the BP-AdaBoost model achieved an R² value of 0.77, compared to 0.73 for the BP-Bagging model and only 0.64 for the standard BP model. Variable importance analysis revealed that trap area, sandstone thickness, sedimentary facies type, and distance to faults significantly contribute to HRP. Furthermore, model accuracy is influenced by multiple factors, including the selection and quantification of geological parameters, dataset size and distribution characteristics, and the choice of machine learning algorithm models. In summary, machine learning provides a reliable method for assessing HRP, offering new insights for identifying high-quality exploration blocks and optimizing development strategies. 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

Background/Objectives: Brain tumors arise from abnormal, uncontrolled cell growth due to changes in the DNA. Magnetic Resonance Imaging (MRI) is vital for early diagnosis and treatment planning. Artificial intelligence (AI), especially deep learning, has shown strong potential in assisting radiologists with MRI analysis. However, many brain tumor classification models achieve high accuracy at the cost of large model sizes and slow inference, limiting their practicality for medical edge computing. In this work we introduce a new attention-guided classification model and explore how model parameters can be reduced without significantly impacting accuracy. Methods: We develop a shallow attention-guided convolutional neural network (ANSA_Ensemble) and evaluate its effectiveness using Monte Carlo simulations, ablation studies, cross-dataset generalization, and Grad-CAM-generated heatmaps. Several state-of-the-art model compression techniques are also applied to improve the efficiency of our classification pipeline. The model is evaluated on three open-source brain tumor datasets. Results: The proposed ANSA_Ensemble model achieves a best accuracy of 98.04% and an average accuracy of 96.69 ± 0.64% on the Cheng dataset, 95.16 ± 0.33% on the Bhuvaji dataset, and 95.20 ± 0.40% on the Sherif dataset. Conclusions: The performance of the proposed model is comparable to state-of-the-art methods. We find that the best tradeoff between accuracy and speed-up factor is consistently achieved using depthwise separable convolutions. The ablation study confirms the effectiveness of the introduced attention blocks and shows that model accuracy improves as the number of attention blocks increases. Our code is made publicly available. Full article