Search Results (1,233)

Modern grid digitalization connects communication networks, monitoring terminals, service platforms, security devices, and operational data sources. Intrusion detection in this setting requires correlating heterogeneous security data with grid-side contextual evidence. To address class imbalance and cross-domain heterogeneity, this study proposes a heterogeneous feature group integration framework for intrusion detection with grid cybersecurity data. Four semantic feature subspaces are constructed symmetrically: network behaviour, power operation context, zone-derived communication/event topology, and system operation state, ensuring equal structural footing for subsequent modality-specific encoding. Transformer-based encoders model temporal dependencies in network, physical, and system state modalities, while a graph neural network encodes topology-related structural information. The resulting embeddings are integrated by a late fusion classifier for multiclass attack identification; the fusion process treats each feature group symmetrically at the decision level, without imposing a priori dominance among modalities. In the main run, the full model achieves an accuracy of 0.944, a macro F1 score of 0.891, a weighted F1 score of 0.937, a macro precision of 0.929, and a macro recall of 0.878. The corresponding balanced accuracy is 0.878, and the multiclass MCC is 0.924. Class-wise results show reliable performance on Benign, Scan, WebAtk, DDoS, DoS, and Backdoor classes, while Ransomware remains difficult and is frequently confused with WebAtk. Specifically, the Ransomware recall is 0.27, with most errors assigned to WebAtk. Modality analysis further indicates that modality contribution is class dependent: some feature groups have limited standalone discriminative power but provide complementary evidence after fusion. This finding highlights an inherent asymmetry in class-wise utility, which we counterbalance by employing both macro and weighted metrics, offering a symmetric evaluation lens that accounts for both minority and majority classes. These results show that grid-oriented intrusion detection benefits from decision-level integration of heterogeneous feature groups and imbalance-aware evaluation, where symmetric treatment of feature subspaces and evaluation perspectives jointly enhances robustness. Full article

Brain lateralization is a biological asymmetry in which a bilaterally organized nervous system develops direction-specific functional organization. This hypothesis distinguishes gravity-driven physical symmetry reduction from informational symmetry breaking. Gravity provides a stable vertical reference, yet matched leftward and rightward tilts become biologically relevant only when noisy vestibular population responses carry decodable tilt-sign information. At fixed unsigned tilt magnitude, the criterion is nonzero conditional mutual information between binary tilt sign and vestibular population response; for equal sign priors, this is equivalent to Jensen–Shannon divergence between sign-conditioned response distributions. Shannon entropy describes within-condition response spread, Fisher information describes local continuous-angle precision, and noise-aware representational distance describes PIVC-centered state separation. The otolith-to-perceptual pathway is formulated as a constrained effective state-space transformation from vestibular population responses through an intermediate brainstem–cerebellar state to distributed parieto-insular vestibular cortex (PIVC)-centered cortical states and perceived self-orientation. The framework predicts sign-specific vestibular and PIVC information for matched tilts, reduced or reorganized sign information in bilateral vestibulopathy, and covariance among cortical geometry, orientation-estimation reliability, and orientation-dependent behavior. Auditory and visual spatial transformations provide computational precedents rather than anatomical homology. The model offers a testable sensorimotor scaffold without determining a fixed hemispheric sign. Full article

Drainage solution reuse in soilless strawberry production can reduce nutrient-rich discharge, but adoption requires microbial control, hydraulic reliability, and manageable energy demand. This field study evaluated a photovoltaic (PV) assisted ultraviolet-C (UV-C) treatment loop for substrate derived drainage solution in a 132 m² three-tier natural light greenhouse producing ‘Solhyang’ strawberry in Sokcho-si, Republic of Korea. The system used an 11.25 kWp vertical windbreak-type PV facility and a 650 W treatment loop comprising a 250 W low-pressure mercury UV-C reactor, a 350 W pump, and a 50 W controller. The loop operated for 2.5 h day⁻¹, processed 3.25 m³ day⁻¹ as cumulative reactor throughput, and consumed 1.625 kWh day⁻¹, equal to 4.22% of the measured daily PV alternating current (AC) output (38.5 kWh day⁻¹). The drainage solution had low ultraviolet transmittance at 254 nm (UVT₂₅₄; 25–50%) and moderate turbidity (5–30 NTU), conditions that can attenuate UV radiation and shield microorganisms. Across six post fruit set sampling events, the mean log₁₀ reductions were 1.15 ± 0.09 for culturable molds/fungal propagules and 1.64 ± 0.09 for culturable aerobic bacteria; paired tests on log₁₀ transformed counts were significant (p < 0.001). Total coliform bacteria were not detected after treatment, corresponding to a detection limit-based lower-bound reduction of ≥2.69 ± 0.17 log₁₀. Apparent fluence values were treated as engineering estimates rather than validated delivered dose. The results support UV-C sanitation as a preliminary enabling step for drainage solution reuse, while biodosimetry, untreated circulation controls, multi-stage seasonal sampling, full-season recirculation, and crop response validation remain necessary. Full article

Green innovation performance (GIP) in new energy enterprises emerges from complex interactions among technological, organizational, and institutional resource subsystems, yet existing research predominantly applies linear, single-factor approaches that fail to capture this systemic complexity. Drawing on the Resource-Based View (RBV) and systems thinking, this study employs fuzzy-set qualitative comparative analysis (fsQCA) on a sample of 54 Chinese A-share listed new energy enterprises—spanning wind power, solar power, hydrogen energy, energy storage, and new energy equipment manufacturing—observed over the 2019–2023 period, to examine the configurational pathways through which these firms achieve high GIP. Green patent grants serve as the outcome measure, and six conditions spanning three resource subsystems are considered: digital transformation and R&D intensity (technological subsystem), firm size and ownership structure (organizational subsystem), and government subsidies and carbon emission performance (institutional subsystem). Three key findings emerge. First, none of the six conditions is individually necessary for high GIP (all consistency scores below 0.90), indicating that high GIP reflects combinations of resources rather than a single driver. Second, the six sufficient configurations identified collapse into two distinct pathway clusters: a “SOE digital-empowerment-driven” cluster, in which digital transformation combines with R&D investment, government subsidies, or organizational scale within state-owned enterprises, and a “resource–capability synergy and substitution” cluster, in which scale resources, R&D investment, and policy support combine with or substitute for digital transformation regardless of ownership. Third, digital transformation appears in five of the six pathways, indicating that it functions as a key—but not universal—enabling element whose effectiveness depends on its alignment with other system components. Beyond confirming that multiple, equally valid resource combinations lead to high GIP, this study’s principal contribution is to embed RBV within an explicit systems framework, showing how technological, organizational, and institutional resources interact as subsystems of a single socio-technical system, and to translate the resulting configurations into differentiated, pathway-specific guidance for enterprises and policymakers navigating the low-carbon energy transition. Full article

This study examines the technical and operational efficiency of publicly listed energy companies operating in Europe, the United States, and Canada during the 2017–2024 energy transition period. The sample includes both traditional electricity utilities and renewable energy producers. Technical efficiency was estimated using output-oriented Data Envelopment Analysis (DEA), specifically the Charnes–Cooper–Rhodes (CCR) and Banker–Charnes–Cooper (BCC) models. Panel-data models were subsequently applied to identify the financial, organisational, regional, and environmental, social, and governance (ESG) factors associated with firm-level efficiency. The results indicate a moderate average level of technical efficiency, with a substantial share of inefficiency attributable to an inappropriate operating scale. Contrary to the initial hypothesis, renewable energy companies were, on average, less technically efficient than traditional utilities, despite achieving higher ESG and environmental scores. European companies exhibited higher efficiency than firms located in the United States and Canada, suggesting that long-term exposure to climate-policy and regulatory pressures may encourage more effective resource use. The panel-model results did not provide robust evidence that ESG performance directly improves technical efficiency. By contrast, profitability, leverage, and firm size were significantly associated with efficiency outcomes. These findings show that the energy transition depends on more than the expansion of renewable energy capacity. Effective resource allocation, financial resilience, organisational adjustment, and an appropriate operating scale are equally important. The study provides relevant implications for corporate managers, investors, and policymakers involved in energy-sector transformation. Full article

The era of highly effective pre-exposure prophylaxis (PrEP) has transformed HIV prevention, yet global HIV incidence remains unacceptably high, particularly in sub-Saharan Africa. While antiretroviral-based prevention is critical, persistent structural barriers to access and adherence highlight the urgent need for a complementary preventive HIV vaccine. In this narrative review and perspective, we argue that scientific progress toward an HIV vaccine must be matched by equally ambitious preparedness investments across three domains: manufacturing and supply chains, clinical trial and regulatory infrastructure, and delivery systems and community engagement. Drawing on lessons from the global rollout of PrEP, post-exposure prophylaxis (PEP), and the COVID-19 pandemic response, we outline a five-pillar strategic roadmap. This roadmap focuses on ensuring that a future HIV vaccine reaches the people who need it most from the moment of regulatory authorization, calling for co-investment in diverse platforms, binding advance market commitments, logistical simulation, and proactive planning for post-licensure effectiveness trials. Preparedness is not secondary to science; it is its essential partner. Full article

Large renewable energy bases are increasingly planned under delivery-hour targets, corridor-capacity constraints, and high shares of wind and photovoltaic generation. Chronological planning samples used in expansion and adequacy studies therefore need to retain not only average renewable-load patterns but also low-probability days with high residual balancing demand, large ramps, curtailment pressure, and sustained renewable scarcity. This paper proposes a risk-anchored order-preserving scenario-chain construction method for renewable energy base planning. The proposed method transforms aligned hourly load, wind, photovoltaic, delivery-demand, and loss-adjusted demand trajectories into distinct operational stress indicators, normalizes them into a joint extreme score, anchors the highest-risk natural days together with their adjacent transition days, and applies clustering only to the remaining regular days. Observed medoid days are then inserted back into chronological order to form a compact scenario chain with explicit weights and adjacency information. A representative 8760 h renewable-base case with 4000 MW wind, 5500 MW photovoltaic, 5400 MW coal support, 1200 MWh storage-energy capacity, a 7600 MW delivery corridor, and a 5600 h delivery target is used to run weight-sensitivity tests and same-budget comparisons against monthly typical days, k-means, k-medoids, hierarchical clustering, Carpe Diem, and seasonal time-series aggregation baselines. Under the equal-weight base case, the proposed chain gives an active-metric mean capture ratio of 1.0660, compared with 0.8377 for monthly typical days. Across four non-equal priority-weight vectors, the active-metric mean remains between 1.0041 and 1.0662. Under the same 151-day budget, conventional k-means, k-medoids, hierarchical clustering, Carpe Diem, and seasonal time-series aggregation baselines obtain active-metric means of 0.8643, 0.9025, 0.8732, 0.9239, and 0.9189, respectively. The results indicate that chronological risk anchoring provides a compact yet physically interpretable sampling layer for planning models that must balance renewable utilization, delivery reliability, and storage adequacy. Full article

This study develops a secondary-data-driven decision support framework for prioritizing strategic energy investment readiness across countries. The empirical application covers 36 countries and 18 criteria grouped under macroeconomic feasibility, institutional capacity, energy security, sustainability and decarbonization, market and demand conditions, and technical resource potential. The study responds to a methodological gap in energy investment prioritization by moving from expert-only linguistic scoring toward a reproducible, explainable, stakeholder-sensitive, and validation-oriented multi-criteria decision support structure. In the implemented empirical model, public secondary data are transformed into a normalized country-level decision matrix, baseline readiness scores are calculated using equal weights, and entropy, Criteria Importance Through Intercriteria Correlation (CRITIC), and hybrid entropy-CRITIC configurations are used as objective weighting benchmarks. Large language model (LLM)-assisted extraction is used as a documented criterion-discovery and screening aid, not as an autonomous scoring, weighting, or ranking mechanism. The fuzzy component is specified as an uncertainty-sensitive extension and illustrated through panel-based fuzzy interval construction for selected time-series indicators, while the main baseline ranking is based on the latest available normalized data matrix. The results show meaningful cross-country variation, with top-10 readiness scores ranging from 0.675 to 0.537 and the lowest five scores ranging from 0.364 to 0.338. Persona-based results indicate that country priorities differ across public planners, private investors, grid operators, sustainability policymakers, and infrastructure funds, while robustness checks show broad ranking stability, with Spearman correlations between 0.892 and 0.986 and a median simulated-agent correlation of 0.932. Fairness and partial-correlation diagnostics suggest that the model captures multidimensional readiness rather than simply reproducing wealth or existing renewable capacity. Post-hoc validation against observable investment- and transition-related benchmarks further supports the convergent validity of the readiness index. The framework should therefore be interpreted as an early-stage country-level screening tool for strategic energy investment prioritization, not as a final project-level investment appraisal model. Full article

To address the challenge of accurately assessing transformer health under multi-source heterogeneous data conditions and to enable precise full-lifecycle lifetime prediction along with interpretable decision-making insights, we propose an ensemble learning framework that integrates multi-model feature selection with Stacking-Optuna optimization. First, we construct a full-lifecycle data framework that integrates a basic data layer with a derived feature layer, resulting in an 89-dimensional feature set. Second, feature importance is evaluated via equal-weighted integration of Random Forest, LightGBM, and XGBoost, and 13 key features are selected using an 80% cumulative importance threshold. On this basis, a Stacking model is constructed using CatBoost, XGBoost, and LightGBM as base learners, with Ridge regression as the meta-learner, while Optuna optimizes the model’s parameters and architecture. Finally, SHAP is employed to perform multi-level interpretability analysis. Experimental results show that the proposed model achieves MAE = 1.3411, RMSE = 1.6899, and R² = 0.9502 on the test set, outperforming the best single model (CatBoost). SHAP analysis further reveals that commissioning year is the feature with the highest SHAP contribution (44.54%), and that feature contributions exhibit a nonlinear pattern, initially decreasing and subsequently increasing with service time. This method achieves high accuracy and interpretability, providing a reference for transformer condition assessment and operational decision-making. Full article

The widespread adoption of voice-based authentication systems has been accompanied by an escalating threat from deep learning-based synthetic speech generation techniques. This study presents a comparative and experimental investigation of cepstral feature extraction methods for deepfake speech detection. Specifically, Mel-Frequency Cepstral Coefficients (MFCC), Linear-Frequency Cepstral Coefficients (LFCC), and Constant-Q Cepstral Coefficients (CQCC) are systematically evaluated with respect to their frequency scaling characteristics, spectral resolution properties, and capacity to capture artifacts specific to synthetic speech production. Experiments were conducted on 5571 audio samples drawn from the ASVspoof 2021 Logical Access evaluation partition, with all methods assessed under identical classification conditions using a linear Support Vector Machine. Results indicate that CQCC attains the highest numerical performance, achieving 83.59% accuracy, 89.15% ROC-AUC, and 15.83% Equal Error Rate (EER); however, the performance difference between MFCC and CQCC does not reach statistical significance (p = 0.202). Five-fold cross-validation corroborates this finding (CQCC: 87.89% ± 0.81%). McNemar’s test confirms that the performance difference between LFCC and CQCC is statistically significant (p = 0.036). A fine-grained attack-wise analysis across 13 spoofing systems reveals that no single feature representation consistently outperforms the others across all attack types; CQCC achieves the highest accuracy on 6 out of 13 systems, while MFCC remains competitive on several attack categories. The overall findings indicate that deepfake detection performance is highly sensitive not only to the classifier architecture but also to the choice of frequency scale, cepstral transformation design, and data conditions. Empirical motivation is provided that multi-feature strategies integrating complementary frequency representations may offer more robust and generalizable detection solutions. Full article

The presence of openings in spherical pressure hulls compromises the structural integrity and induces stress concentrations at the opening area. The strength assessment of shell openings remains a challenge in shell strength theory, even if using an approximated algorithm. While finite element analysis (FEA) is commonly employed, it is computationally prohibitive for iterative design. To address this gap, this study proposes a novel theoretical framework to calculate the Stress Concentration Factor (SCF) for spherical hulls with complex opening reinforcements under external pressure. A geometric transformation method based on the Equal Area Criterion (EAC) is established to convert complex reinforcement geometries into analytical forms. The accuracy is validated through comprehensive FEA and full-scale hydrostatic pressure tests on titanium-alloy manned hulls. The results demonstrate that the proposed method has high accuracy and exhibits excellent applicability, with deviations from experimental data generally within 5%. This study establishes a rapid assessment framework with high accuracy, providing a valuable engineering tool for deep-sea structural design. Full article

Forecasting multi-asset volatility spillovers and turning the forecasts into risk-aware portfolios requires methods that uncover directional information flow between assets, compress the state into a minimal sufficient representation, deliver calibrated uncertainty, and respect explicit tail-risk limits. We propose TDV (Transfer-entropy, Dynamic-graph-attention, Variational-information-bottleneck), an information-theoretic artificial intelligence framework that couples a time-varying transfer entropy network with a graph attention encoder regularised by a variational information bottleneck, and demonstrates the practical value of the calibrated predictive distribution through a downstream entropy-regulated, CVaR-constrained portfolio application. We establish three theoretical results: $L^2$ consistency of the k-nearest-neighbour transfer entropy estimator on $\alpha$-mixing returns with rate $O_P\left(n^{-2/(2+d)}\right)$, a PAC–Bayes generalisation bound of order $O\left(\sqrt{\left(I\right(X;Z)+log(1/\delta \left)\right)/n}\right)$ for the bottleneck-encoded forecaster, and asymptotic CVaR feasibility of the plug-in allocation. In simulations across sparse Granger networks, contagion DCC–GARCH ensembles, and regime-switching factor models, the framework cuts spillover forecasting errors by 24 to 42 percent against LSTM, vanilla GAT, and Transformer baselines, and it recovers 1.6 additional nats of mutual information with the realised connectedness matrix. On a 32-asset global panel covering 2014 to 2025, the model delivers an out-of-sample $R^2$ of 0.331, an annualised Sharpe ratio of 1.46 against 0.83 for an equally weighted benchmark, a maximum drawdown of 7.8 percent, and 95 percent CVaR reductions of 28 to 36 percent across sub-periods relative to a shrinkage minimum-variance baseline. Full article

Calls for transforming biological curricula have emphasized a need for improving quantitative skill development in STEM education. To address this, we designed six interdisciplinary modules to develop quantitative reasoning competencies for a sophomore-level Cell Biology course. After a comprehensive curriculum alignment procedure between a four-year institution and its primary community college sending institutions, we determined module topics, then developed and implemented the modules. We assessed the effects of the modules on student proficiencies using validated pre-post measurements of specific quantitative competencies. Students showed significant total growth in quantitative goals for all modules and for each module individually, even though modules varied widely in difficulty. Transfer students were equally able as direct entry students to gain in quantitative proficiency across the modules, which is an improvement over the findings of a previous study. Additionally, both transfer and direct entry students exposed to more modules had a higher score on a global assessment of quantitative and biological concepts. Attitude assessments showed that students had an overall positive experience with the modules. Our results suggest that adding quantitative modules to core biology courses can promote student understanding of quantitative concepts for both direct entry and transfer students and can benefit transfer students in particular. Full article

To support offline process-parameter screening for Czochralski (CZ) silicon single crystal growth, this paper proposes a surrogate-assisted optimization framework based on an improved Langevin evolutionary algorithm. First, a multi-variable constrained optimization model is established, with the LSA-Transformer-predicted solid–liquid interface deformation used as the objective evaluation and with process-smoothness and physical-feasibility constraints considered. Six key process parameters–heater power, pulling rate, argon flow rate, crystal rotation speed, crucible rotation speed, and magnetic field strength–are selected as decision variables. Second, building on the classical Langevin algorithm, an adaptive inertia weight mechanism, a diversity promoter (DP) operator, and a local escaping operator (LEO) are introduced to improve global exploration and local optima escape in complex search spaces. Verification on 23 classical benchmark functions indicates that the ILEE algorithm shows competitive overall performance and achieves better or comparable results on many functions when compared with particle swarm optimization (PSO), grey wolf optimization (GWO), the original Langevin evolutionary algorithm (LEE), and other baseline algorithms. The proposed framework is then used for offline candidate recipe generation during the crystal equal-diameter growth stage (200 mm, 400 mm, 600 mm, 800 mm, and 1000 mm). The optimized candidate parameter combinations yield lower surrogate-predicted interface deformation under the given LSA-Transformer model and physical constraints. Because these values are not independent CFD or experimental measurements, the results should be interpreted as process-parameter guidance for future physical validation. This work provides a feasible surrogate-assisted offline screening framework for CZ silicon single crystal growth. Full article

This qualitative study explores perceptions of gender equity and leadership in three higher education institutions in Eswatini. The research involved nine senior management members—deans, registrars, and bursars—and eighteen lecturers (nine women and nine men). Employing narrative inquiry, data were gathered through semi-structured interviews and focus groups, capturing rich individual stories and social dynamics. The study explores how gender influences access to leadership roles, the barriers faced, and potential strategies for fostering inclusive environments. Guided by social role theory, the analysis was deductive, examining how cultural norms and stereotypes shape perceptions of leadership and reinforce gender disparities. Women occupy only 22% of senior management roles, while among lecturers, women constitute 50% of the workforce but only 30% of leadership positions, illustrating persistent under-representation. Findings reveal that despite existing policies, cultural norms rooted in hegemonic masculinity continue to impede gender equity, with organisational biases and societal stereotypes maintaining male dominance in leadership. Participants emphasised that policies must be actively enforced, and cultural change initiatives are essential to challenge stereotypes and reshape societal narratives about gender roles. The study underscores the importance of institutional support, mentorship programmes, and visibility initiatives to empower women and promote gender-inclusive leadership. Engaging men as allies is also critical in transforming organisational culture. These findings contribute to advancing understanding of gender dynamics in Eswatini’s higher education sector and highlight the need for comprehensive, context-specific interventions. Addressing these challenges requires coordinated efforts involving policy enforcement, cultural transformation, capacity building, and ongoing evaluation to ensure sustainable progress toward gender equality in academic leadership. Full article