Search Results (135)

This study investigates the use of interpretable machine learning techniques to identify behavioral and demographic patterns associated with diabetes, based on structured population survey data from the Canadian Community Health Survey (CCHS). A decision tree classifier was applied to a dataset comprising $16,824$ respondents and 38 preprocessed features covering lifestyle, well-being, and sociodemographic factors. The model was optimized through grid search with five-fold stratified cross-validation, achieving a test accuracy of $61.3$% (mean $62.6$% ±$0.6$% across a $10\times 5$ repeated stratified cross-validation). Feature importance analysis revealed that age, alcohol consumption patterns, daily energy expenditure, and physical activity were the most influential factors associated with diabetes status, with the top three features exhibiting stable importance across all cross-validation folds. The model produced a set of 32 human-readable decision rules; a sensitivity analysis confirmed that these rules are stable across encoding choices and cross-validation folds. Several model variants were evaluated: a class-weighted decision tree, a logistic regression baseline, an age-only decision tree, and an age and sex logistic regression. The class-weighted model improved minority-class recall (from $0.25$ to $0.53$) at the cost of overall accuracy. A one-hot encoding sensitivity analysis showed that replacing ordinal label encoding of nominal variables with one-hot encoding produces virtually identical results (accuracy: $61.4$% vs. $61.3$%), confirming that the main rules are not artifacts of the encoding choice. Although the classification accuracy is moderate and not significantly better than a majority-class baseline (McNemar’s test, $p=0.455$), the extracted rules confirmed several known associations and revealed interactions between social and lifestyle variables. These rules are intended as hypothesis-generating population-level descriptors rather than validated clinical decision tools, and no causal inference is claimed. This approach demonstrates the value of rule-based models for exploratory public health research. Full article

Scheduling in multi-layer industrial networks must remain stable even when the feedback mechanism of the environment changes inside a single production episode. The system can switch between a step-continuous regime with dense process feedback and a task-driven regime with sparse milestone feedback, so that the same state requires different behaviour before and after the switch. A regime-oblivious policy may therefore optimise the wrong action preference after a switch. We formulate this setting as a mode-switched multi-industrial-chain Markov decision process (MS-MIC-MDP) and prove that a single fixed action preference is necessarily suboptimal in at least one regime. We then propose BHERA, a belief-guided homeostatic estimation framework for regime adaptation. BHERA builds cross-layer representations, performs structured variational inference of slow and fast latent beliefs, estimates the posterior probability of the task-driven regime, and uses that posterior to regulate sample weights, entropy strength, return-prediction emphasis, and latent information capacity. A homeostatic feedback rule on the Kullback–Leibler (KL) divergence keeps the latent representation informative without allowing uncontrolled information growth, and we analyse it as a two-timescale stochastic approximation with an associated convergence argument and a per-iteration complexity bound. Experiments in a multi-layer industrial scheduling simulator show that BHERA achieves higher return, lower cost, and higher utility than CReSCENT, HiTAC-MuSE, Informed Switching, and WToE across all tested perturbations, with paired statistical tests confirming significance. Expanded ablations and parameter-sensitivity studies confirm the importance of regime belief, regime-balanced weighting, bootstrap prediction, homeostatic capacity control, and the dual-timescale latent split. Full article

High-skilled labor plays an important role in regional economic development, yet accurately identifying its influencing patterns remains challenging due to complex factor interactions, spatial spillover effects, and fuzzy boundaries among urban characteristics. Traditional regression-based approaches primarily focus on isolated linear effects, making it difficult to capture multi-factor combinatorial relationships underlying talent agglomeration. To address these limitations, this study proposes a spatially aware fuzzy association rule mining framework by integrating soft-gated spatial weighting and concept stability theory. Using data from the Sixth and Seventh National Population Censuses and the China City Statistical Yearbook, the framework is applied to the Beijing–Tianjin–Hebei (BTH), Yangtze River Delta (YRD), and Middle Reaches of the Yangtze River (MRYR) regions from 2010 to 2020. The results show that the associative patterns of high-skilled labor evolved substantially across regions. In the BTH region, dominant factors shifted from administrative hierarchy and environmental amenities to stronger interactions between economic growth and talent inflow. In the YRD region, economic dynamism gradually replaced static geographic advantages, while in the MRYR region, market-oriented drivers increasingly surpassed administrative-led resource concentration. Overall, the findings suggest a transition from single-factor dependence to multi-factor coupled patterns in China’s regional talent agglomeration. Full article

Background: Metabolic dysfunction-associated steatotic liver disease (MASLD) is a significant and escalating global health concern, with an estimated prevalence of 30%. Current assessments of hepatic steatosis, a hallmark of MASLD, rely on semi-quantitative grading by pathologists, which is inherently limited by inter-observer variability. Objective: To address this limitation, we developed a novel deep learning pipeline, named SteatoStat, to standardize and enhance the quantification of hepatic steatosis in patients with MASLD. Method: The SteatoStat pipeline employs and integrates multiple components such as file format standardization, rule-based cell filtering, and multiple segmentation models across various liver structures, resulting in an output of a continuous quantitative measure of steatosis percentage and translated into steatosis grades. Results: We report a high degree of accuracy and reliability with SteatoStat achieving the following performance metrics (DICE score = 0.8955, AUROC = 0.9928, F1 score = 0.8990). When benchmarked against expert pathologists, the weighted Kappa coefficient is 0.837. Furthermore, in comparison with an existing, well-established model, SteatoStat demonstrated a weighted Kappa coefficient = 0.765. Conclusions: These robust findings underscore its potential clinical utility in providing a standardized objective quantification of hepatic steatosis. Future directions include enhancing the model’s generalizability and its clinical integration through validation on independent, multi-institutional datasets. Full article

The intestinal microbiome is increasingly recognized as an important determinant of canine gastrointestinal health. However, interpreting microbiome sequencing data remains challenging because most analytical approaches rely on taxonomic descriptions, alpha diversity indices, or dysbiosis indices derived generally from a limited number of microbial ecological interpretation targets. While shotgun metagenomic approaches increasingly allow the identification of microbial communities, such analyses remain costly and are not yet widely accessible in routine veterinary settings. The objective of this study was to develop an integrative interpretation framework based on widely accessible biomarkers combining fecal calprotectin and 16S rRNA gene sequencing data. These data enabled the generation of complementary ecological dimensions of gut microbiome organization: biological inflammation assessed through fecal calprotectin, microbiological inflammatory pressure estimated through a Microbiological Inflammatory Score (MIS), and microbiome stability measured by a Microbiome Resilience Score (MRS) derived from alpha diversity, functional balance, and dominance structure. Fecal microbiome profiles obtained by 16S rRNA gene sequencing were analyzed in a real-life cohort of privately owned dogs. Alpha diversity, taxonomic weighting, abundance-dependent dominance rules, beta diversity based on Bray–Curtis dissimilarity, distance to a reference microbiome core, and a 16S-derived dysbiosis score were integrated into a multidimensional interpretation model. Strong ecological associations were observed between resilience, microbial diversity, and dysbiosis-related metrics. Microbiome resilience strongly correlated with Shannon diversity (Spearman ρ = 0.98, p < 0.001), while the reconstructed 16S-derived dysbiosis score showed a more moderate positive correlation with MIS (Spearman ρ = 0.41, p = 0.004), supporting the partially independent ecological dimensions captured by the framework. The results revealed a continuum ranging from stable microbiomes to inflammatory dysbiosis. Most dogs clustered near a reference microbiome core characterized by low microbiological inflammatory pressure and high resilience, whereas a subset of microbiomes showed elevated MIS values, reduced resilience, increased compositional distance from the reference core, and higher dysbiosis index values. These findings support the value of a multidimensional experimental framework integrating inflammation, dysbiosis, and resilience to improve interpretation of canine microbiome profiles under real-life conditions. Full article

Parkinson’s disease is strongly linked to lysosomal dysfunction, particularly reduced activity of glucocerebrosidase (GCase) encoded by the GBA1 gene. Stabilizing GCase using small-molecule modulators represents a promising therapeutic strategy. In this study, phytochemicals from Xylia xylocarpa (Roxb.) Taub., a medicinal plant with reported neuroprotective potential, were profiled using LC-QTOF-MS and evaluated as GCase stabilizers through an integrated computational approach. LC-MS analysis in positive and negative modes tentatively identified 19 metabolites, of which 13 low-molecular-weight compounds (<500 Da) were selected for molecular docking against human GCase. Docking revealed six compounds with higher predicted binding affinity than the reference activator Pyrrolopyrazine. Pharmacokinetic screening based on Lipinski’s rule of five and ADMET predictions identified Senbusine A as a viable lead candidate. It exhibited favorable binding interactions, forming stabilizing contacts within a non-catalytic inter-monomer interface associated with structural modulation of GCase. PASS analysis suggested a high probability of neuroactive properties. Molecular dynamics simulations (200 ns) confirmed stable binding and reduced conformational fluctuations compared to apo and control systems. Overall, computational predictions identify Senbusine A as a potential pharmacological chaperone-like stabilizer of GCase, exhibiting a favorable pharmacological profile and warranting further experimental validation. Full article

Recurrence-Aware Long-Term Cognitive Network (rLTCN) classifiers have reported comparable performance to mainstream black-box models, including tree ensembles and support vector machines, in tabular pattern classification tasks. These classifiers use a two-step learning algorithm to address issues that arise during the training of recurrent neural networks. While the weights in the recurrent block are computed using unsupervised learning, recurrence-aware weights are determined using a one-step learning rule based on the Moore-Penrose inverse. However, the related least-squares learning problem tends to favor easy instances and common patterns, particularly those associated with the majority class in imbalanced datasets. In such scenarios, a loss function that directly optimizes a robust metric, such as the F1 score, would lead to models with stronger generalization capabilities. Unfortunately, incorporating such a metric into the Moore-Penrose inverse learning procedure presents challenges from a mathematical viewpoint. In this paper, we propose four gradient-based correction methods that modify the output logits of rLTCN classifiers once the two-step training process is done. Inspired by procedures such as Platt or Beta scaling, the proposed post-optimization correction methods seek to maximize the F1 score rather than produce calibrated probabilities. The simulations using real-world datasets show that adding a correction layer to rLTCNs improves their performance significantly at the expense of occasional reductions in the precision metric. Full article

Association rule mining is a fundamental data mining technique for uncovering latent relationships among variables in large-scale datasets. However, conventional approaches rely on single-metric filtering strategies, which are insufficient for capturing the inherent multi-criteria nature of rule quality. To address this limitation, this study formulates ARM as a many-objective optimization problem and proposes a hybrid algorithm, NSGA-II/DE-ARM, that simultaneously optimizes four rule-quality measures: support, confidence, lift, and NetConf. The proposed algorithm enhances the NSGA-II framework by integrating binary differential evolution operators, an adaptive operator selection mechanism, lift-weighted tournament selection, and a constraint-domination principle combined with a dynamic minimum support threshold. Its performance was evaluated using two datasets: a SIPRI–World Bank panel dataset consisting of defense industry and macroeconomic indicators covering 46 items over the 2002–2023 period, and the UCI Mushroom benchmark dataset consisting of 118 items. Across 30 independent runs on the SIPRI–World Bank dataset, NSGA-II/DE-ARM outperformed the Apriori baseline in all four metrics (mean lift = 4.748, confidence = 0.853, support = 0.146, NetConf = 0.789), with large effect sizes (Cohen’s d = 1.77–5.77, p < 0.001 in each case). On the Mushroom benchmark dataset, the proposed method also achieved substantial improvements, with Cohen’s d values ranging from 0.93 to 6.16. NSGA-II/DE-ARM generated 68 Pareto-optimal rules in a representative run and achieved the highest hypervolume values on both datasets, with HV = 3.231 for SIPRI–World Bank and HV = 6.262 for Mushroom. These results suggest that NSGA-II/DE-ARM offers decision-makers a broader and more balanced multi-criteria solution set than single-metric filtering approaches. Full article

In the contemporary field of geographic information, place name services serve as a core application support in geographic information science, widely applied in public services, cultural tourism, emergency management, and other scenarios. Place name service composition is a critical link in the integration of spatiotemporal knowledge and intelligent services for place names, determining the ability to rapidly solve complex place name problems. Traditional case-based reasoning methods are primarily rule-driven, making it difficult to deeply integrate semantic and graph structural features, and they also lack precision in measuring the similarity of multi-type place name service cases. To address this, this paper integrates knowledge graphs and case-based reasoning to propose a place name service composition method that balances semantic and graph structural similarity, aiming to enhance the response efficiency and recognition accuracy of complex natural language queries. The method consists of two steps: the first is constructing a knowledge graph case base. Semantic feature extraction is performed on the standard geographic question-answering standard dataset GeoQuery corpus to build a place name service knowledge graph case base that integrates semantic associations and spatial attributes. The second step is constructing a similarity model. The method combines four similarity measures—DeBERTa, TF-IDF, SimHash, and maximum common subgraph—and employs the Analytic Hierarchy Process for weighting to develop a novel similarity evaluation model for case-based reasoning. Experiments demonstrate that this method achieves a 21% improvement in F1-score compared to traditional rule-based methods. Furthermore, the developed prototype system for the intelligent recommendation of place name service composition achieves a recommendation accuracy of 92.64%. This research holds significant practical implications and application value for advancing the geographic information field toward intelligent and precision-based development. Full article

Background/Objectives: Metabolic dysfunction-associated steatotic liver disease (MASLD) is now the leading cause of chronic liver disease globally, mirroring the increasing prevalence of obesity, insulin resistance, and type 2 diabetes. Early detection of hepatic steatosis is vital for cardiometabolic risk assessment; however, conventional imaging is costly and impractical for population screening. This study aimed to develop interpretable machine-learning models to predict ultrasound-detected hepatic steatosis within the MASLD spectrum using routinely available clinical and biochemical data. Methods: We analyzed data from 644 adults, 50% of whom had ultrasound-detected hepatic steatosis. Preprocessing, imputation, and feature selection were implemented within a single scikit-learn pipeline to avoid information leakage. An Elastic Net-regularized logistic regression identified the top 20 predictors, which were subsequently used across nine supervised machine learning (ML) classifiers. Model performance was evaluated via repeated stratified 5-fold cross-validation (25 resamples) using accuracy, F1 score, sensitivity, specificity, Youden’s J, balanced accuracy, and Area Under the Receiver Operating Characteristic Curve (AUROC). Interpretability was assessed using SHapley Additive exPlanations (SHAP). Results: Participants with ultrasound-detected hepatic steatosis exhibited greater adiposity, insulin resistance, and dyslipidemia compared with controls [p < 0.05 for body mass index (BMI), waist circumference, glucose, glycated hemoglobin (HbA1c), triglycerides]. Elastic Net selection highlighted Weight, Ponderal Index, Fibrosis-4 Index (FIB-4), blood urea nitrogen (BUN)/Creatinine ratio, Aspartate Aminotransferase to Platelet Ratio Index (APRI), and Visceral Adiposity Index as the strongest predictors. Logistic Regression and Gradient Boosting achieved the best performance (accuracy = 0.65 ± 0.03; AUROC = 0.71 ± 0.04; balanced accuracy = 0.66 ± 0.06), outperforming rule-based indices such as Fatty Liver Index (FLI) and Hepatic Steatosis Index (HSI) reported in the literature. SHAP analysis confirmed clinically coherent feature effects, with higher anthropometric and hepatic injury indices increasing the predicted probability of ultrasound-detected hepatic steatosis. Conclusions: Routinely available clinical and biochemical parameters can predict hepatic steatosis with moderate accuracy using transparent, interpretable ML models. Logistic Regression and Gradient Boosting provided best discrimination and robust internal performance, offering a pragmatic, low-cost approach for early identification of ultrasound-detected hepatic steatosis within the MASLD spectrum in primary and metabolic care settings. Full article

Nonlinear state estimation plays an important role in aerospace sensing applications, where estimation accuracy must be balanced against computational efficiency. In this paper, a sparse-grid Gaussian kernel quadrature Kalman filter (SGKQKF) is proposed for discrete-time nonlinear state estimation by combining Gaussian kernel quadrature (GKQ) weighting with a Smolyak sparse-grid construction. The univariate GKQ rule is constructed on scaled Gauss–Hermite nodes through a truncated Mercer eigendecomposition of the Gaussian kernel and is then extended to multivariate cases via the Smolyak construction to alleviate the curse of dimensionality associated with tensor-product rules. The proposed method is positioned within the established sparse-grid filtering framework, with the specific contribution of integrating kernel-adapted quadrature weights into sparse-grid structures for discrete-time nonlinear Gaussian filtering. For fixed nodes, the exact kernel-quadrature weights minimize the worst-case integration error in the reproducing kernel Hilbert space (RKHS) induced by the Gaussian kernel, whereas the closed-form weights used in the implementation are interpreted as a Mercer-based practical approximation to this exact rule, with the approximation error characterized through the Mercer spectral-tail expression of the Gaussian kernel. For sparse grids, where a closed-form RKHS optimality result is not available, numerical maximum mean discrepancy (MMD) evaluations are presented as empirical diagnostics in the tested configurations. Numerical experiments demonstrate that the proposed filter achieves a favorable accuracy–efficiency trade-off compared with conventional deterministic Gaussian filters. Full article

To effectively alleviate the common problems of information noise and information loss in personalized recommendation systems, as well as to address data sparsity and cold-start issues, this paper proposes a collaborative filtering recommendation model that integrates user attributes and association rules, named Impoved_UARCF. The model introduces a user attribute-sensitive module and a user-item rating-sensitive module to perform deep feature modeling from the perspectives of multi-dimensional user attributes and user-item rating interactions, respectively. The user attribute-sensitive module employs a similarity computation mechanism based on user attributes to mine and decouple deep attribute features among users, enhancing the discriminability and generalization ability of feature representations, thereby effectively resolving information noise and information loss. The user-item rating-sensitive module utilizes association rule mining technology to learn the relationship weights between users in real time, enabling accurate aggregation and propagation of user-item rating features, thus effectively addressing data sparsity and cold-start problems. Extensive experiments conducted on three public datasets verify the superiority of Impoved_UARCF in recommendation performance, as well as the effectiveness, scalability, and robustness of each module design. Full article

Background: Hospitalized patients with chronic kidney disease (CKD) are at high risk for acute kidney injury (AKI), dialysis, and mortality, yet CKD is often treated as a clinically homogeneous condition. Whether distinct cardiometabolic comorbidity patterns define meaningful inpatient CKD subgroups with differential outcome risks remains unclear. Methods: We conducted a retrospective cross-sectional study of adult hospitalizations for CKD using the 2022 Healthcare Cost and Utilization Project National Inpatient Sample. Hospitalizations were classified into five mutually exclusive CKD phenotypes using a rule-based framework based on diabetes mellitus, heart failure, hypertension, and vascular disease: isolated, hypertensive/vascular, metabolic, cardiorenal, and multimorbid cardiometabolic. Outcomes included AKI, dialysis during hospitalization, and in-hospital mortality. Survey-weighted multivariable logistic regression models were used to estimate adjusted odds ratios (aORs). Sensitivity analyses excluded end-stage kidney disease and dialysis dependence and restricted this study to non-transfer hospitalizations. The effect modification by age was assessed for dialysis. Results: Among 1,062,813 CKD hospitalizations, the unadjusted outcome rates varied substantially across phenotypes. After adjustment, cardiorenal CKD was associated with higher odds of acute kidney injury (aOR 1.16, 95% CI 1.12–1.19) and in-hospital mortality (aOR 1.54, 95% CI 1.50–1.58), whereas multimorbid cardiometabolic CKD demonstrated the strongest association with dialysis during hospitalization (aOR 2.34, 95% CI 2.25–2.43). Hypertensive/vascular CKD was not associated with a difference in mortality risk, while metabolic CKD was associated with a lower adjusted mortality rate compared to isolated CKD. Integrated analyses revealed distinct phenotype-specific risk profiles rather than a single severity gradient. Our findings were robust across the sensitivity analyses, and age significantly modified phenotype–dialysis associations. Conclusions: Hospitalized CKD populations exhibit marked phenotype-specific heterogeneity in AKI, dialysis, and mortality risk. A simple, clinically interpretable phenotype framework identifies distinct inpatient failure patterns and may inform future studies evaluating phenotype-specific risk stratification and management strategies. Full article

China’s rural public open spaces (POS) are largely governed as common-pool resources through self-organized collective arrangements, often regarded as a viable pathway to sustainable commons management. Yet, in practice, these systems remain prone to overuse and under-maintenance, reflecting collective action failures associated with the tragedy of the commons. The governance of rural POS therefore constitutes a complex social–ecological problem shaped by the interplay of institutional rules, biophysical conditions, and user–stakeholder interactions. Taking Taigu District in Shanxi Province—characterized by heterogeneous social–ecological contexts and collective action dilemmas—as the empirical case, this study develops a meso-level baseline model to identify the key conditions (design principles) for sustainable rural POS governance. Adopting an expert-based epistemological approach, 24 specialists in rural governance (scholars, planners, and local administrators) were engaged. Grounded in commons and collective action theories within the Social–Ecological Systems (SES) framework and informed by Transaction Cost Economics (TCE), the study operationalizes a Delphic Hierarchy Process (DHP), combining three rounds of Delphi to establish consensus on governance conditions with the Analytic Hierarchy Process (AHP) to derive their relative weights. The model specifies 14 governance conditions across four interrelated dimensions: ecological (e.g., clearly defined resource boundaries and congruence between resource characteristics and user needs), institutional (e.g., simple and enforceable rules, accessible conflict-resolution mechanisms, accountable monitoring, and calibrated external support), social (e.g., social capital, leadership capacity, clearly defined user boundaries, and group interdependence), and interactional (e.g., resource dependence, equity in benefit distribution, and supply–demand alignment). It further clarifies their relative importance and systemic interdependencies. By operationalizing commons design principles within a meso-level analytical framework, the study advances their empirical application in rural planning and offers five targeted managerial implications to strengthen institutional robustness and the long-term sustainability of self-governed rural POS. Full article

This study re-examines the monetary “black box” in India by investigating whether monetary-policy transmission is state-dependent across different interest-rate environments. Using quarterly data spanning 1993Q1–2024Q2, it constructs a Taylor rule-based monetary-policy shock to mitigate the endogeneity of raw policy rates and estimates dynamic discrete-threshold regressions with endogenously determined regimes. The results provide strong evidence of nonlinearity and structural instability in India’s transmission process. For real output, the weighted average call money rate (WACR) emerges as the more informative threshold variable, while wholesale price inflation is more effectively segmented by the 91-day Treasury bill yield. The findings show that the contractionary effect of monetary policy on output is most evident in the intermediate-rate regime, whereas low- and high-rate regimes exhibit weaker or counterintuitive short-run responses, consistent with crisis accommodation, delayed pass-through, and state-specific frictions. For inflation, monetary tightening is associated with a short-run price puzzle in low- and intermediate-yield regimes but produces the expected disinflationary effect in the high-yield regime. Across channels, the credit and asset-price channels matter selectively for output, while the exchange-rate channel is the most relevant for inflation only in the intermediate regime. Overall, the evidence suggests that monetary-policy transmission in India is regime-dependent and that policy assessment should distinguish between operating-rate conditions and broader market-rate regimes. Full article