Search Results (1,002)

High-quality test cases are vital for ensuring software reliability and security. However, existing symbolic execution tools generally rely on single-path search strategies, have limited feature extraction capability, and exhibit unstable model predictions. These limitations make them prone to local optima in complex or cross-scenario tasks and hinder their ability to balance testing quality with execution efficiency. To address these challenges, this paper proposes a Deep Active Ensemble Learning Framework for symbolic execution path exploration. During training, the framework integrates active learning with ensemble learning to reduce annotation costs and improve model robustness, while constructing a heterogeneous model pool to leverage complementary model strengths. In the testing stage, a dynamic ensemble mechanism based on sample similarity adaptively selects the optimal predictive model to guide symbolic path exploration. In addition, a gated graph neural network is employed to extract structural and semantic features from the control flow graph, improving program behavior understanding. To balance efficiency and coverage, a dynamic sliding window mechanism based on branch density enables real-time window adjustment under path complexity awareness. Experimental results on multiple real-world benchmark programs show that the proposed framework detects up to 16 vulnerabilities and achieves a cumulative 27.5% increase in discovered execution paths in hybrid fuzzing. Furthermore, the dynamic sliding window mechanism raises the F1 score to 93%. Full article

Concrete compressive strength is a fundamental indicator of the mechanical properties of High-Performance Concrete (HPC) with multiple components. Traditionally, it is measured through laboratory tests, which are time-consuming and resource-intensive. Therefore, this study develops a machine learning-based regression framework to predict compressive strength, aiming to reduce experimental costs and resource usage. Under three different data preprocessing strategies—raw data, standard score, and Box–Cox transformation—a selected set of high-performance ensemble models demonstrates excellent predictive capacity, with both the coefficient of determination (${\mathrm{R}}^2$) and explained variance score (EVS) exceeding 90% across all datasets, indicating high accuracy in compressive strength prediction. In particular, stacking ensemble (${\mathrm{R}}^2$-$0.920$, $\mathrm{EVS}$-$0.920$), XGBoost regression (${\mathrm{R}}^2$-$0.920$, $\mathrm{EVS}$-$0.920$), and HistGradientBoosting regression (${\mathrm{R}}^2$-$0.913$, $\mathrm{EVS}$-$0.914$) based on Box–Cox transformation data show strong generalization capability and stability. Additionally, tree-based and boosting methods demonstrate high effectiveness in capturing complex feature interactions. Furthermore, this study presents an analytical workflow that enhances feature interpretability through visualization techniques—including Partial Dependence Plots (PDP), Individual Conditional Expectation (ICE), and SHapley Additive exPlanations (SHAP). These methods clarify the contribution of each feature and quantify the direction and magnitude of its impact on predictions. Overall, this approach supports automated concrete quality control, optimized mixture proportioning, and more sustainable construction practices. Full article

Accurate fault diagnosis of power transformers is critical for maintaining grid reliability, yet conventional dissolved gas analysis (DGA) methods face challenges in feature representation and high-dimensional data processing. This paper presents an intelligent diagnostic framework that synergistically integrates systematic feature engineering, tensor decomposition-based feature selection, and a sparrow search algorithm (SSA)-optimized multi-kernel support vector machine (MKSVM) for transformer fault classification. The proposed approach first expands the original five-dimensional gas concentration measurements to a twelve-dimensional feature space by incorporating domain-driven IEC 60599 ratio indicators and statistical aggregation descriptors, effectively capturing nonlinear interactions among gas components. Subsequently, a novel Tucker decomposition framework is developed to construct a three-way tensor encoding sample–feature–class relationships, where feature importance is quantified through both discriminative power and structural significance in low-rank representations, successfully reducing dimensionality from twelve to seven critical features while retaining 95% of discriminative information. The multi-kernel SVM architecture combines radial basis function, polynomial, and sigmoid kernels with optimized weights and hyperparameters configured through SSA’s hierarchical producer–scrounger search mechanism. Experimental validation on DGA samples across seven fault categories demonstrates that the proposed method achieves 98.33% classification accuracy, significantly outperforming existing methods, including kernel PCA-based approaches, deep learning models, and ensemble techniques. The framework establishes a reliable and accurate solution for transformer condition monitoring in power systems. Full article

Road traffic crashes are a major global challenge, resulting in significant loss of life, economic burden, and societal impact. This study seeks to enhance the precision of traffic accident prediction using advanced machine learning techniques. This study employs an ensemble learning approach combining the Random Forest, the Bagging Classifier (Bootstrap Aggregating), the Extreme Gradient Boosting (XGBoost) and the Light Gradient Boosting Machine (LightGBM) algorithms. To address class imbalance and feature relevance, we implement feature selection using the Extra Trees Classifier and oversampling using the Synthetic Minority Over-sampling Technique (SMOTE). Rigorous hyperparameter tuning is applied to optimize model performance. Our results show that the ensemble approach, coupled with hyperparameter optimization, significantly improves prediction accuracy. This research contributes to the development of more effective road safety strategies and can help to reduce the number of road accidents. Full article

While most patients with degenerative rotator cuff tears respond to conservative treatment, a minority progress to surgery. To anticipate these cases under class imbalance, we propose a sensitivity-constrained evolutionary feature selection framework prioritizing surgical-class recall, benchmarked against traditional methods. Two variants are proposed: (i) a single-objective search maximizing balanced accuracy and (ii) a multi-objective search also minimizing the number of selected features. Both enforce a minimum-sensitivity constraint on the minority class to limit false negatives. The dataset includes 347 patients (66 surgical, 19%) described by 28 clinical, imaging, symptom, and functional variables. We compare against 62 widely adopted pipelines, including oversampling, undersampling, hybrid resampling, cost-sensitive classifiers, and imbalance-aware ensembles. The main metric is balanced accuracy, with surgical-class F1-score as secondary. Pairwise Wilcoxon tests with a win–loss ranking assessed statistical significance. Evolutionary models rank among the top; the multi-objective variant with a Balanced Bagging Classifier performs best, achieving a mean balanced accuracy of 0.741. Selected subsets recurrently include age, tear location/severity, comorbidities, and pain/functional scores, matching clinical expectations. The constraint preserved minority-class recall without discarding or synthesizing data. Sensitivity-constrained evolutionary feature selection thus offers a data-preserving, interpretable solution for pre-surgical decision support, improving balanced performance and supporting safer triage decisions. Full article

Chlorophyll content is a key physiological indicator reflecting photosynthetic capacity, and the Soil–Plant Analysis Development (SPAD) meter is a commonly used tool for its rapid and non-destructive estimation. Hyperspectral imaging (HSI) is a non-destructive technique that captures fine spectral characteristics and thus holds great potential for high-throughput phenotyping and early stress detection. This study aimed to explore the potential of HSI combined with ensemble learning (EL) to estimate SPAD of rapeseed seedlings under different durations of waterlogging. Hyperspectral images and corresponding SPAD values were collected from six rapeseed cultivars at 0, 2, 4 and 6 days of waterlogging. The mutual information was employed to select the top 30 most relevant spectral and vegetation index features. The EL model was constructed using partial least squares, support vector machine, random forest, ridge regression and elastic net as the first-layer learners and a multiple linear regression as the second-layer learner. The results showed that the EL model showed superior stability and higher prediction accuracy compared to single models across various genotypes and waterlogging treatment datasets. As waterlogging duration increased, the overall model accuracy improved; notably, under 6 days of waterlogging, the EL model achieved an R² of 0.79 and an RMSE of 3.27, indicating strong predictive capability. This study demonstrated that combining EL with HSI enables stable and accurate estimation of SPAD values, therefore providing an effective approach for early stress monitoring in crops. Full article

Global climate change has led to frequent extreme weather events such as high temperatures and droughts, severely threatening the heat and water balance during the growing season of summer maize. To adapt to these changes, adjusting planting dates to optimize crop development has become a key agronomic measure for mitigating climate stress and ensuring yield. Against this backdrop, precise monitoring of leaf area index (LAI) is crucial for evaluating the effectiveness of planting date regulation and achieving precision management. To reveal the impact of planting date variations on summer maize LAI inversion and address the limitations of single data sources in comprehensively reflecting complex environmental conditions affecting crop growth, this study examined summer maize at different planting dates across the North China Plain. Through stepwise regression analysis (SRA), multiple vegetation indices (VIs) and 0–2nd order fractional order derivatives (FODs), spectral parameters were dynamically screened. These were then integrated with effective accumulated temperature (EAT) to optimize model inputs. Partial Least Squares Regression (PLSR), Random Forest (RF), Support Vector Regression (SVR), and Adaptive Boosting Regression (AdaBoot) algorithms were employed to construct LAI inversion models for summer maize across different planting dates and mixed planting dates. Results indicate that, compared to empirical VIs and “tri-band” parameters, randomly selected dual-band combination VIs exhibit the strongest correlation with summer maize LAI. Key bands identified through SRA screening concentrated in the 0.7–1.2 order range, primarily distributed across the red edge and near-infrared bands. Multi-feature models incorporating EAT significantly improved retrieval accuracy compared to single-feature models. Optimal models and feature combinations varied across planting dates. Overall, the VIs + EAT combination exhibited the highest stability across all models. Ensemble learning algorithms RF and AdaBoost performed exceptionally well, achieving average R² values of 0.93 and 0.92, respectively. The model accuracy for the 20-day delayed planting (S4) decreased significantly, with an average R² of 0.62, while the average R² for other planting dates exceeded 0.90. This indicates that the altered environmental conditions during the later growth stages of LAI due to delayed planting hindered LAI estimation. This study provides an effective method for estimating summer maize LAI across different planting dates under climate change, offering scientific basis for optimizing adaptive cultivation strategies for maize in the North China Plain. Full article

Accurate prediction of hemolytic peptides is essential for peptide safety evaluation and therapeutic design; however, existing models remain constrained by limited accuracy and interpretability. To overcome these challenges, we propose a regression framework that integrates embeddings from a protein language model with handcrafted amino acid descriptors. Specifically, sequence representations derived from the ESM2_t33 model are fused with physicochemical amino acid descriptor features, and key predictive variables are selected through a three-stage strategy involving variance filtering, F-test ranking, and mutual information analysis. The final ensemble model, composed of Random Forest, Extremely Randomized Trees, Gradient Boosting, eXtreme Gradient Boosting (XGBoost), and Ridge Regression, achieved a coefficient of determination (R²) of 0.57 and a correlation coefficient (R) of 0.76 on the test set, outperforming previous approaches. To enhance interpretability, we applied Shapley value analysis and the Calibrated_Explanation algorithm to quantify feature contributions and generate reliable sample-specific explanations. The trained model has been deployed online as HemPepPred, a tool for predicting hemolytic concentration (HC₅₀) values, which provides a practical platform for rational peptide design and safety assessment. Full article

The kiln head temperature of a rotary kiln is a core process parameter in cement clinker production, and its accurate prediction coupled with uncertainty quantification is crucial for process optimization, energy consumption control, and safe operation. To tackle the prediction challenges arising from strong multi-variable coupling and nonlinear time series characteristics, this paper proposes a prediction approach integrating feature selection, heterogeneous model ensemble, and probabilistic interval estimation. Firstly, the Relief algorithm is adopted to select key features and construct a time series feature set with high discriminability. Then, a hierarchical architecture encompassing deep feature extraction, heterogeneous model fusion, and probabilistic interval quantification is devised. CNN is utilized to extract spatial correlation features among multiple variables, while BiLSTM is employed to bidirectionally capture the long-term and short-term temporal dependencies of the temperature sequence, thereby forming a deep temporal–spatial feature representation. Subsequently, RF is introduced to establish a heterogeneous model ensemble mechanism, and dynamic weight allocation is implemented based on the Mean Absolute Error of the validation set to enhance the modeling capability for nonlinear coupling relationships. Finally, Gaussian probabilistic regression is leveraged to generate multi-confidence prediction intervals for quantifying prediction uncertainty. Experiments on the real rotary kiln dataset demonstrate that the R² of the proposed model is improved by up to 15.5% compared with single CNN, BiLSTM and RF models, and the Mean Absolute Error is reduced by up to 27.7%, which indicates that the model exhibits strong robustness to the dynamic operating conditions of the rotary kiln and provides both accuracy guarantee and risk quantification basis for process decision-making. This method offers a new paradigm integrating feature selection, adaptive heterogeneous model collaboration, and uncertainty quantification for industrial multi-variable nonlinear time series prediction, and its hierarchical modeling concept is valuable for the intelligent perception of complex process industrial parameters. Full article

The coexistence of class imbalance and overlap poses a major challenge in classification and significantly limits model accuracy. Data-level methods alleviate class imbalance by generating samples, but without ensuring their rationality, which may introduce noise. Algorithm-level methods are designed based on the model training process, avoiding noise introduction. However, existing methods often fail to consider the potential multiclass scenarios within overlap regions or design targeted solutions for different overlap patterns. This paper proposes an ensemble imbalanced classification framework via dual-perspective overlapping analysis with multi-resolution metrics. The dataset is divided into multiple resolutions for independent analysis, capturing distributional information from local to global levels. For each independent resolution, overlap is analyzed from the perspectives of “feature overlap” and “instance overlap” to derive more refined overlap scores. Flow model mapping and importance weighting are, respectively, applied to refine overlapping samples according to the two criteria. During testing, classifiers are adaptively selected based on the overlap degree of test samples under different criteria, and predictions across resolutions are integrated for the final decision. Experiments on 39 datasets demonstrate that the proposed method outperforms typical imbalanced classification methods in F-measure and G-mean, with particularly notable gains on 15 severely overlapping datasets. Full article

In recent years, smart grid security has gained considerable attention. Numerous studies have proposed techniques to detect cyber-attacks using sensor data; however, limited attention has been given to distinguishing cyber intrusions from physical faults in the power grid. In this paper, we present a supervised intrusion–disturbance classification pipeline to accurately differentiate physical faults from cyber-attacks. First, we augment raw channels with relation-centered features to emphasize relative contrasts and suppress common-mode effects, then we apply embedded feature selection via LightGBM to retain a compact, informative subset. Class imbalance is addressed through class weighting, and an Extremely Randomized Trees classifier serves as the core learner. Experiments on 15 datasets cover both binary (Attack vs. Natural) and multiclass (Attack/Natural/NoEvents) regimes. The approach attains 98.44% mean accuracy for the binary task and 98.22% for the multiclass task, demonstrating consistent discrimination between cyber-attacks, physical faults, and normal operation. The results indicate that relational features combined with embedded selection and a tree ensemble offer a practical, accurate alternative to heavier deep models for smart-grid monitoring. Full article

To address the issue of weak geoelectric current field signals that are severely affected by noise and cannot be directly used for geological structure analysis in injected geoelectric current field detection technology, this study proposes a complete ensemble empirical mode decomposition with adaptive noise and improved wavelet thresholding collaborative denoising (CEEMDAN-IWT) method to enhance the interpretation accuracy of geoelectric current signals. The method performs signal decomposition through CEEMDAN and selects the effective intrinsic mode function (IMF) components based on the variance contribution criterion for preliminary denoising. It then combines the improved wavelet thresholding function for further fine denoising and reconstruction, obtaining high signal-to-noise ratio (SNR) electrical data. Simulation and real-world data validation show that in a simulation experiment with an initial SNR of −5 dB, the method improves the $SNR$ to 18.65 dB, and the SNR enhancement is superior to traditional methods under various noise intensities. In practical applications, the normalized cross-correlation (NCC) between the denoised signal and the original injected signal reaches as high as 0.9254, significantly outperforming traditional methods. At the same time, it balances the preservation of signal features with noise suppression, offering significant application value for improving the reliability of injected geoelectric current field detection data. Full article

Forest canopy height is essential for ecosystem process modeling and carbon stock assessment. However, most prediction approaches rely on sparse or interpolated LiDAR labels, leading to uncertainties in heterogeneous forests where laser footprints are limited or unevenly distributed. To address these issues, this study proposes a multi-source ensemble learning framework that uses airborne PolInSAR-derived continuous canopy height as training labels for accurate forest height prediction. The framework features two key innovations: (1) a hybrid baseline selection strategy (PROD+ECC) within the PolInSAR inversion, significantly improving the quality and stability of initial labels; (2) a dual-layer ensemble learning model that integrates machine learning and deep learning to interpret multi-source features (Landsat-8, GEDI, DEM, and kNDVI), enabling robust upscaling from local inversion to regional prediction. Independent validation in Gabon’s Akanda National Park achieved R² = 0.748 and RMSE = 5.873 m, reducing RMSE by 43.6% compared with existing global products. This framework mitigates sparse supervision and extrapolation bias, providing a scalable paradigm for high-accuracy canopy height mapping in complex tropical forests and offering an effective alternative to LiDAR-based approaches for global carbon assessment. Full article

Cortisol and testosterone are key digital biomarkers reflecting neuroendocrine activity across the hypothalamic–pituitary–adrenal (HPA) and hypothalamic–pituitary–gonadal (HPG) axes, encoding stress adaptation and behavioral regulation. Continuous real-world monitoring remains challenging due to the sparsity of sensing and the complexity of multimodal data. This study introduces a synthetic sensor-driven computational framework that models hormone variability through data-driven simulation and predictive learning, eliminating the need for continuous biosensor input. A hybrid deep ensemble integrates biological, behavioral, and contextual data using bidirectional multitask learning with one-dimensional convolutional neural network (1D-CNN) and long short-term memory (LSTM) branches, meta-gated expert fusion, Bayesian variational layers with Monte Carlo Dropout, and adversarial debiasing. Synthetically derived longitudinal hormone profiles that were validated by Kolmogorov–Smirnov (KS), Wasserstein, maximum mean discrepancy (MMD), and dynamic time warping (DTW) metrics account for class imbalance and temporal sparsity. Our framework achieved up to 99.99% macro F1-score on augmented samples and more than 97% for unseen data with ECE below 0.001. Selective prediction further maximized the convergence of predictions for low-confidence cases, achieving 99.9992–99.9998% accuracy on 99.5% of samples, which were smaller than 5 MB in size so that they can be employed in real time when mounted on wearable devices. Explainability investigations revealed the most important features on both the physiological and behavioral levels, demonstrating framework capabilities for adaptive clinical or organizational stress monitoring. Full article

With the rapid growth of photovoltaic (PV) installed capacity, accurate prediction of PV power is crucial for the safe and flexible operation of power grids. However, PV output sequences exhibit strong non-stationarity and a superposition of high-frequency disturbances and low-frequency trends, resulting in multi-frequency aliasing. Traditional models struggle to capture both long-term dependencies and short-term details, while multi-task learning (MTL) often suffers from negative transfer, limiting prediction accuracy. This paper proposes a hybrid PV power forecasting framework integrating complementary ensemble empirical mode decomposition with adaptive noise (CEEMDAN), PatchTST reconstruction, and progressive layered extraction (PLE) MTL. First, conventional models tend to prioritize learning low-frequency features while ignoring weak high-frequency signals under multi-frequency aliasing, which cannot meet the requirement for precise frequency-sensitive PV power prediction. To address this problem, CEEMDAN is employed to decompose the PV sequence into intrinsic mode functions (IMFs). Next, the fluctuation complexity of each IMF is quantified via RCMSE and K-means clustering: high-frequency components are captured using small patches to preserve details, while low-frequency components use larger patches to learn long-term trends. Subsequently, a PatchTST-BiLSTM reconstruction network with patch partitioning and multi-head attention is adopted to capture temporal dependencies and optimize data representation, overcoming the bottleneck caused by the imbalance between long-term and short-term features. Finally, recursive feature elimination (RFE) feature selection combined with a PLE multi-task network can coordinate expert models to mitigate negative transfer and enhance high-frequency response capability. Experiments on the Alice Springs dataset show that the proposed method significantly outperforms conventional deep learning and new multi-task models in the mean absolute error (MAE) and root mean square error (RMSE). The results show that, compared with the MTL_Attention_LSTM method, the proposed method reduces the average MAE by 45.9% and RMSE by 44.6%, achieving more accurate forecasting of PV power. Full article