Search Results (87)

Gross Primary Productivity (GPP) plays a vital role in the terrestrial carbon cycle and ecosystem functioning. Understanding its spatio-temporal dynamics and driving mechanisms is critical for predicting ecosystem responses to climate change. China’s GPP has experienced complex responses due to heterogeneous climate, environment, and human activities, yet their impacts and interactions across ecosystems remain unquantified. This study used the Mann–Kendall test and SHapley Additive exPlanations to quantify the contributions and interactions of climate, vegetation, topography, and human factors using GPP data (2001–2020). Nationally, GPP showed a significant upward trend, particularly in deciduous broadleaf forests, croplands, grasslands, and savannas. Leaf area index (LAI) is identified as the primary contributor to GPP variations, while climate factors exhibit nonlinear interactive effects on the modeled GPP. Ecosystem-specific sensitivities were evident: forest GPP is predominantly associated with climate–vegetation coupling. Additionally, in coniferous forests, the interaction between anthropogenic factors and topography shows a notable association with productivity patterns. Grassland GPP is primarily linked to topography, while cropland GPP is mainly related to management practices and environmental conditions. In contrast, the GPP of savannas and shrublands is less influenced by factor interactions. These findings high-light the necessity of ecosystem-specific management and restoration strategies and provide a basis for improving carbon cycle modeling and climate change adaptation planning. Full article

The growing reliance on solar energy necessitates robust and interpretable forecasting models for stable grid management. Current research frequently employs Explainable AI (XAI) to glean insights from complex black-box models, yet the reliability and consistency of these explanations remain largely unvalidated. Inconsistent feature attributions can mislead grid operators by incorrectly identifying the dominant drivers of solar generation, thereby affecting operational planning, reserve allocation, and trust in AI-assisted decision-making. This study addresses this critical gap by conducting a systematic statistical evaluation of feature rankings generated by multiple XAI methods, including model-agnostic (SHAP, PDP, PFI, ALE) and model-specific (Split- and Gain-based) techniques, within a time-series regression context. Using a LightGBM model for one-day-ahead solar power forecasting across four sites in Calgary, Canada, we evaluate consensus and divergence using the Friedman test, Kendall’s W, and Spearman’s rank correlation. To ensure the generalizability of our findings, we further validate the results using a CatBoost model. Our results show a strong overall agreement across methods (Kendall’s W: 0.90–0.94), with no statistically significant difference in ranking (p > 0.05). However, pairwise analysis reveals that the “Split” method frequently diverges from other techniques, exhibiting lower correlation scores. These findings suggest that while XAI consensus is high, relying on a single method—particularly the split count—poses risks. We recommend employing multi-method XAI and using agreement as an explicit diagnostic to ensure transparent and reliable solar energy predictions. Full article

Coal and gas outbursts constitute a major hazard for mining safety, which is critical for the sustainable development of China’s energy industry. Rapid, accurate, and reliable pre-diction is pivotal for preventing and controlling outburst incidents. Nevertheless, the mechanisms driving coal and gas outbursts involve highly complex influencing factors. Four main geological indicators were identified by examining the attributes of these factors and their association to outburst intensity. This study developed a machine learning-based prediction model for outburst risk. Five algorithms were evaluated: K Nearest Neighbors (KNN), Back Propagation (BP), Random Forest (RF), Support Vector Machine (SVM), and eXtreme Gradient Boosting (XGBoost). Model optimization was performed via Bayesian hyperparameter (BO) tuning. Model performance was assessed by the Receiver Operating Characteristic (ROC) curve; the optimized XGBoost model demonstrated strong predictive performance. To enhance model transparency and interpretability, the SHapley Additive exPlanations (SHAP) method was implemented. The SHAP analysis identified geological structure was the most important predictive feature, providing a practical decision support tool for mine executives to prevent and control outburst incidents. Full article

Background: Optical coherence tomography (OCT) can quantify the morphology and dimensions of a macular hole for diagnosis and treatment planning. Objective: The aim of this study was to perform automatic segmentation of macular holes (MHs) and cysts from OCT macular volumes using a deep learning-based model and to quantitatively evaluate decision reliability using the model’s focus regions and GradientSHAP-based explainability. Methods: In this study, we automatically segmented MHs and cysts in OCT images from the open-access OIMHS dataset. The dataset comprises 125 eyes from 119 patients and 3859 OCT B-scans. OCT B-scan slices were input to a UNet-48-based model with a 2.5D stacking strategy. Performance was evaluated using Dice and intersection-over-union (IoU), boundary accuracy was evaluated using the 95th-percentile Hausdorff distance (HD95), and calibration was evaluated using the expected calibration error (ECE). Explainability was quantified from GradientSHAP maps using lesion coverage and spatial focus metrics: Attribution Precision in Lesion (APILτ), which is the proportion of attributions (SHAP contributions) falling inside the lesion; Attribution Recall in Lesion (ARILτ), which is the proportion of the true lesion covered by the attributions; and leakage (Leakτ = 1 − APILτ), which is the proportion of attributions falling outside the lesion. Spatial focus was monitored using the center-of-mass distance (COM-dist), which is the Euclidean distance between the attribution center and the segmentation center. All metrics were calculated using the top τ% of the pixels with the highest SHAP values. SHAP features were clustered using PCA and k-means. Explanations were calculated using the clinical mask in ground truth (GT) mode and the model segmentation in prediction (Pred) mode. Results: The Dice/IoU values for holes and cysts were 0.94/0.91 and 0.87/0.81, respectively. Across lesion classes, HD95 = 6 px and ECE = 0.008, indicating good boundary accuracy and calibration. In GT mode (τ = 20), three regimes were observed: (i) retina-dominant: high ARIL (hole: 0.659; cyst: 0.654), high Leak (hole: 0.983; cyst: 0.988), and low COM-dist (hole: 7.84 px; cyst: 6.91 px), with the focus lying within the retina and largely confined to the retinal tissue; (ii) peri-lesional: highest ARIL (hole: 0.684; cyst: 0.719), relatively lower Leak (hole: 0.917; cyst: 0.940), and medium/high COM-dist (hole: 16.22 px; cyst: 10.17 px), with the focus located around the lesion; (iii) narrow-coverage: primarily seen for cysts in GT mode (ARIL: 0.494; Leak: 1.000; COM-dist: 52.02 px), with markedly reduced coverage. In Pred mode, the ARIL20 for holes increased in the retina-dominant cluster (0.758) and COM-dist decreased (6.24 px), indicating better agreement with the model segmentation. Conclusions: The model exhibited high accuracy and good calibration for MH and cyst segmentation in OCT images. Quantitative characterization of SHAP validated the model results. In the clinic, peri-lesion and narrow-coverage conditions are the key situations that require careful interpretation. Full article

Traditional single-prediction models often exhibit limitations in meeting wind power prediction requirements in complex operational scenarios. Furthermore, the inherent “black-box” nature of deep learning models leads to limited interpretability of predictions, hindering effective support for grid dispatch planning. To address these issues, this study proposes a novel day-ahead wind power prediction method, referred to as SHapley Additive exPlanations (SHAP)–Mixture of Experts (MoE), which integrates SHAP into an MoE framework. Here, SHAP is employed for interpretability purposes. This study innovatively transforms SHAP analysis into prior knowledge to guide the decision-making of the MoE gating network and proposes a two-layer dynamic interpretation mechanism based on the collaborative analysis of gating weights and SHAP values. This approach clarifies key meteorological factors and the model’s advantageous scenarios, while quantifying the uncertainty among multiple expert decisions. Firstly, each expert model was pre-trained, and its parameters were frozen to construct a candidate expert pool. Secondly, the SHAP vectors for each pre-trained expert were computed over all sample features to characterize their decision-making logic under varying scenarios. Thirdly, an augmented feature set was constructed by fusing the original meteorological features with SHAP attribution matrices from all experts; this set was used to train the gating network within the MoE framework. Finally, for new input samples, each frozen expert model generates a prediction along with its corresponding SHAP vector, and the gating network aggregates these predictions to produce the final forecast. The proposed method was validated using operational data from an offshore wind farm located in southeastern China. Compared with the best individual expert model and traditional ensemble forecasting models, the proposed method reduces the Root Mean Square Error (RMSE) by 0.23% to 4.92%. Furthermore, the method elucidates the influence of key features on each expert’s decisions, offering insights into how the gating network adaptively selects experts based on the input features and expert-specific characteristics across different scenarios. Full article

Dam break problem-induced floods can trigger hazardous dike overtopping, demanding predictions that are fast, accurate, and interpretable. We pursue two objectives: (i) introducing a new alpha evolution (AE) optimization scheme to improve tree-model predictive accuracy, and (ii) developing a cluster-wise modeling strategy in which regimes are defined by wave characteristics. Using a dataset generated via physical model experiments and smoothed particle hydrodynamics (SPH) numerical simulations, we first group samples via hierarchical clustering (HC) on the Froude number $\left(\mathrm{Fr}\right)$, wave nonlinearity $\left(R\right)$, and relative distance to the dike $\left(D\right)$. We then benchmark CatBoost, XGBoost, and ExtraTrees within each cluster and select the best-performing CatBoost as the baseline, after which we train standard CatBoost and its AE-optimized variant. Under random train–test splits, AE-CatBoost achieves the strongest generalization for predicting relative run-up distance $H_m$ (testing dataset $R^2=0.9803$, $\mathrm{RMSE}=0.0599$), outperforming particle swarm optimization (PSO) and grid search (GS)-tuned CatBoost. We further perform TreeSHAP analyses on AE-CatBoost for global, local, and interaction attributions. SHAP analysis yields physics-consistent explanations: D dominates, followed by H and L, with a weaker positive effect of $\mathrm{Fr}$ and minimal influence of R; $H\times D$ is the strongest interaction pair. Overall, AE optimization combined with HC-based cluster-wise modeling produces accurate, interpretable overtopping predictions and provides a practical route toward field deployment. Full article

The dual imperative of mitigating carbon emissions and maximizing hydrocarbon recovery has amplified global interest in carbon capture, utilization, and storage (CCUS) technologies. These integrated processes hold significant promise for achieving net-zero targets while extending the productive life of mature oil reservoirs. However, their effectiveness hinges on a nuanced understanding of the complex interactions between geological formations, reservoir characteristics, and injection strategies. In this study, a comprehensive machine learning-based framework is presented for estimating CO₂ storage capacity and enhanced oil recovery (EOR) performance simultaneously in subsurface reservoirs. The methodology combines simulation-driven uncertainty quantification with supervised machine learning to develop predictive surrogate models. Simulation results were used to generate a diverse dataset of reservoir and operational parameters, which served as inputs for training and testing three machine learning models: Random Forest, Extreme Gradient Boosting (XGBoost), and Artificial Neural Networks (ANN). The models were trained to predict three key performance indicators (KPIs): cumulative oil production (bbl), oil recovery factor (%), and CO₂ sequestration volume (SCF). All three models exhibited exceptional predictive accuracy, achieving coefficients of determination (R²) greater than 0.999 across both training and testing datasets for all KPIs. Specifically, the Random Forest and XGBoost models consistently outperformed the ANN model in terms of generalization, particularly for CO₂ sequestration volume predictions. These results underscore the robustness and reliability of machine learning models for evaluating and forecasting the performance of CO₂-EOR and sequestration strategies. To enhance model interpretability and support decision-making, SHapley Additive exPlanations (SHAP) analysis was applied. SHAP, grounded in cooperative game theory, offers a model-agnostic approach to feature attribution by assigning an importance value to each input parameter for a given prediction. The SHAP results provided transparent and quantifiable insights into how geological and operational features such as porosity, injection rate, water production rate, pressure, etc., affect key output metrics. Overall, this study demonstrates that integrating machine learning with domain-specific simulation data offers a scalable approach for optimizing CCUS operations. The insights derived from the predictive models and SHAP analysis can inform strategic planning, reduce operational uncertainty, and support more sustainable oilfield development practices. Full article

Understanding outdoor thermal environments at fine spatial scales is essential for developing climate-responsive urban and building design strategies. This study investigates the determinants of local air temperature deviations in Seoul, Korea, using high-resolution in situ sensor data integrated with multi-source urban and building information. Hourly temperature records from 436 road-embedded sensors (March 2024–February 2025) were transformed into relative metrics representing deviations from the network-wide mean and were combined with semantic indicators derived from street-view imagery—Green View Index (GVI), Road View Index (RVI), Building View Index (BVI), Sky View Index (SVI), and Street Enclosure Index (SEI)—along with land-cover and building attributes such as impervious surface area (ISA), gross floor area (GFA), building coverage ratio (BCR), and floor area ratio (FAR). Employing an eXtreme Gradient Boosting (XGBoost)–Shapley Additive exPlanations (SHAP) framework, the study quantifies nonlinear and interactive relationships among morphological, environmental, and visual factors. SEI, BVI, and ISA emerged as dominant contributors to localized heating, while RVI, GVI, and SVI enhanced cooling potential. Seasonal contrasts reveal that built enclosure and vegetation visibility jointly shape micro-scale heat dynamics. The findings demonstrate how high-resolution, observation-based data can guide climate-responsive design strategies and support thermally adaptive urban planning. Full article

Graph learning is well suited to modeling relationships among communicating entities in network intrusion detection. However, the resulting models are frequently difficult to interpret, in contrast to many classical approaches that offer more transparent reasoning. This work integrates SHapley Additive exPlanations with temporal, edge-aware GNN based on GraphSAGE architecture to deliver an explainable, inductive intrusion detection model for NetFlow data named TE-G-SAGE. Using the NF-UNSW-NB15-v3 dataset, flow data are transformed into temporal communication graphs where flows are directed edges and endpoints are nodes. The model learns relational patterns across two-hop neighborhoods and achieves strong recall under chronological evaluation, outperforming a GCN baseline and recovering more attacks than a tuned XGBoost model. SHAP is adapted to graph inputs through a feature attribution on the two-hop computational subgraph, producing global and local explanations that align with analyst reasoning. The resulting attributions identify key discriminative features while revealing shared indicators that explain cross-class confusion. The research shows that temporal validation, inductive graph modeling, and Shapley-based attribution can be combined into a transparent, reproducible intrusion detection framework suited for operational use. Full article

As the foundational carrier of socio-economic development and ecological security, territorial space reflects the degree of coordination between functional structure and efficiency output. However, most existing evaluation methods overlook the heterogeneous functional endowments of spatial units and therefore cannot reasonably assess the efficiency that each unit should achieve under comparable conditions. To address this limitation, this study proposes a function-oriented and interpretable framework for territorial spatial efficiency evaluation based on the Production–Living–Ecological (PLE) paradigm. An entropy-weighted indicator system is constructed to measure production, living, and ecological efficiency, and an XGBoost–SHAP model is developed to infer the nonlinear mapping between functional attributes and efficiency performance and to estimate the ideal efficiency of each spatial unit under Quanzhou’s prevailing macro-environment. By comparing ideal and observed efficiency, functional–efficiency deviations are identified and spatially diagnosed. The results show that territorial efficiency exhibits strong spatial heterogeneity: production and living efficiency concentrate in the southeastern coastal belt, whereas ecological efficiency dominates in the northwestern mountainous region. The mechanisms differ substantially across dimensions. Production efficiency is primarily driven by neighborhood living and productive conditions; living efficiency is dominated by structural inheritance and strengthened by service-related spillovers; and ecological efficiency depends overwhelmingly on local ecological endowments with additional neighborhood synergy. Approximately 45% of spatial units achieve functional–efficiency alignment, while peri-urban transition zones and hilly areas present significant negative deviations. This study advances territorial efficiency research by linking functional structure to efficiency generation through explainable machine learning, providing an interpretable analytical tool and actionable guidance for place-based spatial optimization and high-quality territorial governance. Full article

Background: Artificial intelligence (AI) emerges as a powerful tool to assist breast cancer screening; however, its integration into different mammographic modalities remains insufficiently explored. Digital Mammography (DM) is widely accessible but presents limitations in dense breast tissue, whereas Contrast-Enhanced Spectral Mammography (CESM) provides functional information that enhances lesion visualization. Understanding how deep learning models behave across these modalities, and determining whether their decision-making patterns remain consistent, is essential for equitable clinical adoption. Methods: This study evaluates three convolutional neural network (CNN) architectures, ResNet-18, DenseNet-121, and EfficientNet-B0, for binary classification of breast lesions using DM and CESM images from the public CDD-CESM dataset (2006 images, three diagnostic classes). The models are trained separately on DM and CESM using three classification tasks: Normal vs. Benign, Benign vs. Malignant, and Normal vs. Malignant. A 3-fold cross-validation scheme and an independent test set are employed. Training uses transfer learning with ImageNet weights, weighted binary cross-entropy (BCE) loss, and SHapley Additive exPlanations (SHAP) analysis to visualize pixel-level relevance of model decisions. Results: CESM yields higher performance in the Normal vs. Benign and Benign vs. Malignant tasks, whereas DM achieves the highest discriminative ability in the Normal vs. Malignant comparison (EfficientNet-B0: AUC = 97%, Accuracy = 93.15%), surpassing the corresponding CESM results (AUC = 93%, Accuracy = 85.66%). SHAP attribution maps reveal anatomically coherent decision patterns in both modalities, with CESM producing sharper and more localized relevance regions due to contrast uptake, while DM exhibits broader yet spatially aligned attention. Across architectures, EfficientNet-B0 demonstrates the most stable performance and interpretability. Conclusions: CESM enhances subtle lesion discrimination through functional contrast, whereas DM, despite its simpler acquisition and wider availability, provides highly accurate and explainable outcomes when combined with modern CNNs. The consistent SHAP-based relevance observed across modalities indicates that both preserve clinically meaningful information. To the best of our knowledge, this study is the first to directly compare DM and CESM under identical preprocessing, training, and evaluation conditions using explainable deep learning models. Full article

The rapid growth of the consumer credit card market has introduced substantial regulatory and risk management challenges. To address these challenges, financial institutions increasingly adopt advanced machine learning models to improve default prediction and portfolio monitoring. However, the use of such models raises additional concerns regarding transparency and fairness for both institutions and regulators. In this study, we investigate the consistency of Shapley Additive Explanations (SHAPs), a widely used Explainable Artificial Intelligence (XAI) technique, through a case study on credit card probability-of-default modeling. Using the Default of Credit Card dataset containing 30,000 consumer credit accounts information, we train 100 Extreme Gradient Boosting (XGBoost) models with different random seeds to quantify the consistency of SHAP-based feature attributions. The results show that the feature SHAP stability is strongly associated with feature importance level. Features with high predictive power tend to yield consistent SHAP rankings (Kendall’s W = 0.93 for the top five features), while features with moderate contributions exhibit greater variability (Kendall’s W = 0.34 for six mid-importance features). Based on these findings, we recommend incorporating SHAP stability analysis into model validation procedures and avoiding the use of unstable features in regulatory or customer-facing explanations. We believe these recommendations can help enhance the reliability and accountability of explainable machine learning framework in credit risk management. Full article

Background/Objectives: Differentiated thyroid cancer (DTC) represents over 90% of all hyroid malignancies and typically has a favorable prognosis. However, approximately 30% of patients experience recurrence within 10 years after initial treatment. Conventional risk classification frameworks such as the American Thyroid Association (ATA) and AJCC TNM systems rely heavily on pathological interpretation, which may introduce observer variability and incomplete documentation. This study aimed to develop an interpretable machine-learning framework for risk stratification in DTC and to identify major clinical predictors using SHapley Additive exPlanations (SHAP). Methods: A retrospective dataset of 345 patients was obtained from the UCI Machine Learning Repository. Thirteen clinicopathological features were analyzed, including Age, Gender, T, N, M, Hx Radiotherapy, Focality, Adenopathy, Pathology, and Response. Statistical analysis and feature selection (ReliefF and mRMR) were applied to identify the most influential variables. Two modeling scenarios were tested using an optimizable neural network classifier: (1) all 10 core features and (2) reduced features selected from machine learning criteria. SHAP analysis was used to explain model predictions and determine feature impact for each risk category. Results: Reducing the input features from 10 to 6 led to improved performance in the explainable neural network model (AUC = 0.94, accuracy = 92%), confirming that T, N, Response, Age, M, and Hx Radiotherapy were the most informative predictors. SHAP analysis highlighted N and T as the dominant drivers of high-risk classification, while Response enhanced postoperative biological interpretation. Notably, when Response was excluded (Scenario III), the optimizable tree model still achieved strong predictive performance (AUC = 0.93–0.96), demonstrating that accurate preoperative risk estimation can be achieved using only clinical baseline features. Conclusions: The proposed interpretable neural network model effectively stratifies recurrence risk in DTC while reducing dependence on subjective pathological interpretation. SHAP-based feature attribution enhances clinical transparency, supporting integration of explainable machine learning into thyroid cancer follow-up and personalized management. Full article

Reliable fault diagnosis in industrial robots is essential for minimizing downtime and ensuring safe operations. Conventional model-based methods often require detailed system knowledge and struggle with unmodeled dynamics, while purely data-driven approaches can achieve good accuracy but may not fully exploit the underlying structure of robot motion. In this study, we propose a feature-informed machine learning framework for fault detection in robotic manipulators. A multi-layer perceptron (MLP) is trained to estimate robot dynamics from joint states, and SHapley Additive exPlanations (SHAP) values are computed to derive discriminative feature representations. These attribution patterns, or SHAP fingerprints, serve as enhanced descriptors that enable reliable classification between normal and faulty operating conditions. Experiments were conducted using real-world data collected from industrial robots, covering both motor brake faults and reducer anomalies. The proposed SHAP-informed framework achieved nearly perfect classification performance (0.998 ± 0.003), significantly outperforming baseline classifiers that relied only on raw kinematic features (0.925 ± 0.002). Moreover, the SHAP-derived representations revealed fault-consistent patterns, such as enhanced velocity contributions under frictional effects and joint-specific shifts for reducer faults. The results demonstrate that the proposed method provides high diagnostic accuracy and robust generalization, making it well suited for safety-critical applications and predictive maintenance in industrial robotics. Full article

Precise rainfall-runoff modeling (RRM) is vital for disaster management, resource conservation, and mitigation. Recent deep learning-based methods, such as long short-term memory (LSTM) networks, often struggle with major challenges, including temporal sensitivity, feature selection, generalizability, and hyperparameter tuning. The objective of this study is to develop an accurate and generalizable rainfall–runoff modeling framework that addresses the four aforementioned challenges. We propose a novel RRM framework that integrates transductive LSTM (TLSTM) to capture fine-grained temporal changes, off-policy proximal policy optimization (PPO) combined with Shapley Additive exPlanations (SHAP)-based reward functions for feature selection, an enhanced generative adversarial network (GAN) for online data augmentation, and Bayesian optimization hyperband (BOHB) for efficient hyperparameter tuning. TLSTM uses transductive learning, where samples near the test point are given extra weight, to capture fine-grained temporal shifts. Off-policy PPO contributes to this process by selecting features sensitive to temporal patterns in RRM. Our improved GAN conducts online data augmentation by excluding some gradients, increasing diversity and relevance in synthetic data. Finally, BOHB accelerates hyperparameter tuning by merging Bayesian optimization with the scaling efficiency of Hyperband. We evaluate our model using the Comprehensive Attributes and Meteorology for Large-Sample Studies (CAMELS) dataset under individual and regional scenarios. It achieves Nash–Sutcliffe efficiency (NSE) scores of 0.588 and 0.873, surpassing the baseline scores of 0.548 and 0.830, respectively. The generalizability of our approach was assessed on the hydro-climatic datasets for North America (HYSETS), also yielding improved performance. These improvements indicate more accurate capture of flow dynamics and peak events, supporting a robust and interpretable framework for RRM. Full article