Search Results (4,682)

Multimodal emotion recognition systems based on electroencephalography (EEG) and eye tracking (ET) provide complementary information about neural and visual responses; however, practical deployment requires balancing classification performance and computational cost. This study investigates this trade-off by systematically evaluating 60 feature representation configurations generated from different EEG channel regions, frequency bands, feature types, and signal modalities. Support Vector Machine (SVM), Random Forest (RF), and Artificial Neural Network (ANN) classifiers were evaluated under subject-independent leave-one-subject-out (LOSO) cross-validation, resulting in 180 realizable classifier-feature configuration systems. Median Macro-F1 was used as the primary performance objective, while median classifier inference time was used as the computational-cost objective. A multi-stage selection strategy integrating Δ-based near-optimal filtering and Pareto dominance analysis was applied to identify performance-efficient systems. The highest median Macro-F1 (0.6474) was achieved by an ANN using temporal delta-band PSD features combined with ET information, with a median classifier inference time of 0.0057 s. This system remained the final selected system across Δ values of 0.01, 0.02, and 0.03. An ET-only ANN baseline achieved a median Macro-F1 of 0.6265, indicating a modest improvement when temporal delta-band EEG information was added. These findings demonstrate that feature representation, modality composition, classifier choice, and computational cost should be considered jointly when designing subject-independent multimodal emotion recognition systems. Full article

Green bonds and green bank loans coexist as instruments for financing the low-carbon transition, yet the strategic mechanism through which issuers select between them remains weakly formalized. This paper models green debt instrument choice as a three-stage extensive-form game with perfect information in which a firm first selects a financing route, a bank then chooses credit enhancement in the bond branch or loan terms in the loan branch, and investors decide whether to subscribe. Because the bank’s credit enhancement enters the investors’ participation condition, the three players are genuinely strategically interdependent, and because the bond branch carries only a small standby cost rather than the full loan cost, the regulatory regime affects the choice of instrument and not merely the level of financing cost. The subgame-perfect equilibrium is characterized analytically by backward induction and is then implemented numerically over the eight admissible states of demand, issuer credibility and regulatory regime through a Monte Carlo experiment of 200,000 parameter draws. The share of simulated parameter configurations yielding a green bond equilibrium falls from 93.4% under high demand and strong credibility to 14.2% when both deteriorate; the regulatory regime flips the equilibrium instrument in 19.4% of draws, against 0% in the degenerate special case without the enhancement channel. A variance-based global sensitivity analysis (Sobol and Morris) and an incomplete-information extension with a noisy credibility signal complete the analysis. All reported quantities are numerical solutions of the model under an illustrative calibration, not empirical estimates. Full article

In hyperspectral image classification, the most suitable classifier depends on the magnitude and the type of degradation present in the scene; this information, however, cannot be obtained without labels at the stage where the choice has to be made. A two-layer diagnosis computed before classification and without labels is proposed in this study: a severity index reports the magnitude of the degradation, while three scene-derived indicators separate independent, band-dependent, spectrally correlated and striped structures from one another, while classification itself remains supervised. The approach is evaluated on ten benchmark scenes under a leakage-aware spatial protocol. Classifier fragility is found to follow two opposing regimes: tree ensembles are robust to spatially local degradation but fragile to degradation spread across the spectrum, whereas the opposite is observed for the convolutional network. Regime-based classifier selection produces the better classification map in every combination examined. Replacing the spatial protocol with a random pixel split is further shown to inflate accuracy by between 0.162 and 0.249 and to change the best classifier in six of the ten scenes. Full article

In Bangladesh, adolescent girls face multiple challenges that negatively influence diet quality. Understanding the determinants of their diet quality is essential for informing effective nutrition interventions. This rapid review synthesised evidence on determinants of diet quality among Bangladeshi adolescent girls using the Socio-Ecological Model (SEM). We searched MEDLINE (Ovid), CINAHL Complete and Web of Science. Eligible studies examined determinants of diet quality-related indicators, such as dietary diversity, nutrient intake, or food choices among adolescent girls in Bangladesh. Fifteen studies met the inclusion criteria. We conducted a narrative synthesis. The studies reported the determinants across five SEM levels. At the individual level, adolescent girls’ diet quality was associated with their taste preferences, perceived body image, knowledge of nutrition and health, self-efficacy, food choice motives, dieting concerns, and misconceptions during menstruation. At the interpersonal level, family decision-making dynamics, gender-biased food allocation, household wealth and food security status, expenditure on food, low parental education, family and peer influence, and women’s empowerment were associated with diet quality. At the organisational level, lack of dedicated lunchrooms, availability of unhealthy food near schools, nutrition education by community organisations, and advice from healthcare providers were associated with diet quality. Community-level determinants included cultural norms, rural-urban disparities, and geographic variations. At the policy/macro level, food prices and seasonal food availability emerged as critical determinants. Diet quality among adolescent girls in Bangladesh is shaped by complex multi-level factors spanning individual to policy-level factors. These insights can guide context-appropriate interventions to improve diet quality in this population. Full article

The rapid diffusion of electric vehicles (EVs) is weakening the conventional fuel tax base, creating new challenges for transport-energy taxation. Although South Korea does not currently tax EV charging electricity, such taxation is becoming increasingly relevant as EV adoption expands. This study examines consumer preferences for EV fuel-tax design by focusing on revenue recycling and tax salience, operationalized as the visibility of tax information on EV charging receipts. Using stated-preference data from a discrete choice experiment with South Korean adults aged 20–59, we estimate a mixed-mixed multinomial logit model capturing both discrete preference segments and continuous within-segment heterogeneity. A post-estimation multinomial logit analysis of individual-level salience coefficients identifies characteristics associated with relatively strong positive or negative preferences for tax salience. The results reveal two distinct segments. The majority shows positive preferences for tax salience and the revenue-use outcomes, whereas the minority is highly cost-sensitive and favors direct EV performance improvements. Stronger preferences for tax salience are associated with greater knowledge of the tax system, while negative preferences are associated with lower education and not currently driving. The results therefore represent conditional preferences among alternative EV fuel-tax designs, rather than acceptance of an EV fuel tax relative to no tax. The study provides policy-relevant insights into consumer preferences for EV fuel-tax design during the electric mobility transition. Full article

In an increasingly complex modern social information environment, in which consumers are exposed to both changes in economic conditions and pervasive social comparison cues, there remains limited research exploring the relationship between financial anticipation, social signals, and consumption choices. Using an experimental approach, we examine how income trajectories and social information shape consumption choices. Participants in our online decision task were randomly exposed to information indicating either increasing or decreasing income trajectories, alongside social information about the popularity of a relatively cheap or expensive choice of a specific type of product, among individuals with the same income trajectory. Firstly, we find that cheap information during economic decline undermines the standard prediction that anticipated income decline leads consumers to prefer cheaper products. Specifically, the results suggest a significant non-conformity effect specific to the scenario of anticipated income decline: when income was expected to decrease and the social information indicated a majority choice of cheap products, participants were more likely to select the expensive products. On the other hand, social information had no significant impacts when income was expected to increase. Our findings challenge the conventional intuition that social information indicating a cheap majority encourages frugal consumption, instead suggesting a counter-conformity pattern under anticipated income decline. Full article

The increasing frequency of wildfires has made large-scale collective emergency evacuation increasingly critical. However, existing studies provide limited understanding of how information structures shape the interaction between individual risk perception and collective evacuation behavior. This study develops a collective evolutionary game-based evacuation framework under ambiguous and clear information conditions. Under ambiguous information, individual heterogeneity in risk sensitivity, mobility, and resource endowment is incorporated into social interaction payoffs. Under clear information, observable evacuation consequences, including travel time, risk exposure, and congestion effects derived from route-choice interactions, are incorporated into evacuation utility. Numerical simulations examine the evolutionary characteristics of collective evacuation behavior under different information conditions and population scales. The results show that social interactions play an important role in shaping evacuation decisions under ambiguous information, while congestion effects and route-choice interactions influence evacuation utility under large-scale demand. Sensitivity analyses further demonstrate that congestion representation affects evacuation utility across different population scales. These findings highlight the importance of considering information structure, individual heterogeneity, and collective interactions in evacuation modeling. Emergency management should therefore improve risk communication, evacuation capacity, and congestion mitigation strategies. This study provides theoretical insights into collective evacuation decision-making under heterogeneous information conditions. Full article

Individual tree segmentation and diameter at breast height (DBH) estimation are fundamental to precision forest inventory. Light Detection and Ranging (LiDAR) technology has become an essential tool for achieving these objectives at the single-tree level. Different LiDAR platforms—notably unmanned aerial vehicle (UAV) and Mobile Laser Scanning (MLS)—each offer distinct advantages in capturing forest structural information. Multi-source LiDAR fusion has been proposed as a strategy to combine these complementary strengths. However, how to effectively select segmentation algorithms across different LiDAR data sources remains insufficiently understood, particularly for subtropical coniferous plantations with heterogeneous canopy structure. This study systematically compared four individual tree segmentation algorithms (Donager2021, Dalponte2016, Silva2016, and marker-controlled watershed segmentation [MCWS]) across three LiDAR data sources (UAV-only, MLS-only, and fused UAV–MLS) in Pinus massoniana plantations in subtropical China. DBH estimation models were then developed based on the best-performing segmentation results to examine whether data fusion simultaneously improves both detection and DBH retrieval accuracy. The main findings are as follows: (1) the three canopy height model (CHM)-based algorithms achieved a mean overall accuracy (OA) for individual-tree detection of approximately 64% on UAV data but fell below 30% on MLS data, failing to support effective detection; (2) The Donager2021 algorithm, which directly exploits trunk structure from three-dimensional point clouds, achieved the highest OA of 93.15% with MLS data and further improved to 94.82% with fused data; (3) DBH estimation reached a mean R² of 0.96 for both MLS and fused datasets, yet MLS LiDAR alone produced a lower RMSE (2.31 cm; rRMSE = 5.91%) than fused LiDAR (RMSE = 2.40 cm; rRMSE = 6.24%); and (4) higher detection accuracy did not necessarily lead to better DBH estimation, revealing a trade-off between the two objectives. These findings indicate that fusion does not universally improve all downstream tasks, and that the choice of LiDAR configuration and segmentation algorithm should be guided by specific inventory objectives. Full article

In this study, we use gratifications (U&G) theory to explore how social media shapes Millennials’ gastronomic destination choices. Rather than treating social media influence as uniform, we identify three key gratifications obtained from gastronomy-related content: experiential–inspirational, utilitarian decision-making, and information search and assurance. The study investigates how these gratifications directly and indirectly affect destination choice through pre-visit destination impressions. Data were collected from 418 Chinese Millennials. Validity of measurements was confirmed through exploratory factor analysis (EFA), confirmatory factor analysis (CFA), and reliability tests, with hypotheses tested using Hayes’ PROCESS Model 4. Findings indicate that pre-visit destination impression significantly predicts destination choice and mediates the influence of all three gratifications. Experiential–inspirational gratification has the greatest overall impact; experiential–inspirational and information search and assurance gratifications operate through both directly and indirectly, while utilitarian decision-making gratification operates primarily through pre-visit impressions. By integrating U&G theory with pre-visit impressions, this study offers a novel pathway-specific explanation of social media’s role in gastronomic destination choice, suggesting that different gratifications activate distinct psychological mechanisms rather than a uniform process. It also advances destination-choice research by highlighting pre-visit impression as a rapid evaluative mechanism linking digital gratification and behavioral decision. Full article

Urban CO₂ marginal abatement costs (MACs) provide important information for designing sustainable low-carbon transition strategies, but their interpretation may be affected by reference technology choices and identification uncertainty. Using 5004 city-year observations from 278 Chinese prefecture-level cities over 2006–2023, this study applies a Global non-radial directional distance function to compare National and four-region Group reference technologies under identical baseline settings. At each fixed frontier projection, we characterize the complete admissible range of supporting shadow prices rather than select a single dual solution and assess sensitivity across seven prespecified modeling dimensions. Under the National benchmark, 87.31% of observations are bounded-set identified. National and Group identified sets overlap in 74.34% of city-years, indicating that strict benchmark ordering is uncommon. Across baseline–alternative comparisons, identification status changes in 9.87% of cases, whereas National–Group direction reversals occur in 0.84%; temporal technology generates the largest identification response (25.92%). Baseline numerical diagnostics show successful primal–dual solutions, unique projections, and no nesting violations. The results support more transparent sustainability-oriented assessments of urban decarbonization by reporting MACs together with identification status, benchmark choice, and specification sensitivity rather than as unconditional scalar values. Full article

This study audits whether regime weighting can sharpen conformal forecast intervals without using target-period information or omitting the required finite-sample correction. Conformal prediction builds such intervals from past forecast errors. A natural refinement gives more weight to errors from days whose volatility resembles the forecast day. On daily natural-rubber prices, the refinement appears to work: intervals become about 20% narrower than plain split-conformal, with a significantly better Winkler score. This paper asks whether that gain is real. Three implementation choices are examined, one at a time. The first uses a regime signal that already sees the price move it is meant to predict. The second estimates the regime model on the same residuals the interval is calibrated on. The third omits a correction that the weighted quantile requires in finite samples. The forecast-feasible construction—predictive regime probabilities, a validation-fitted regime model, and the finite-sample correction—shows no detectable improvement over split-conformal, at a paired Winkler difference of $+0.06$ (95% CI $-0.15$ to $+0.17$). Applying the correction alone is not always enough: it removes the apparent advantage on the VMD-augmented ridge residuals, but the filtered comparison arm survives it on the AR(1) residuals at $-0.54$. Only withholding target-period information eliminates the artifact on both. Concentrated weights also leave some intervals unbounded, whereas adaptive conformal baselines remain finite throughout. A controlled simulation reproduces the same apparent gain where no regime information exists at all. Apparent sharpness must therefore be audited for information timing, calibration reuse, and the finite-sample correction. Full article

Background: Dietitians play a crucial role in promoting sustainable dietary patterns aimed at maintaining both human and planetary health. Several studies have investigated knowledge, attitudes, and practices (KAPs) regarding food sustainability among dietitians; nevertheless, evidence from the Middle East remains scarce, with no studies conducted in Saudi Arabia. Objective: This study aimed to assess KAPs regarding sustainable dietary patterns and dietary adherence among dietitians in Saudi Arabia. Methods: An online structured questionnaire was distributed to collect self-reported information on food sustainability, including knowledge and perceived understanding, perceived importance of key concepts, behaviors, and food consumption. A total of 221 dietitians completed the survey. Results: Dietitians reported greater engagement with health-oriented and practical sustainability behaviors, particularly food waste reduction, fruit and vegetable consumption, and affordability-related choices. Also, gaps in dietitians’ knowledge and familiarity were identified regarding environmental metrics of food sustainability concepts. Dietitians reported that food consumption was characterized by frequent consumption of animal-based foods, moderate consumption of fruits, vegetables, legumes, and nuts, and low adoption of plant-based alternatives. Conclusions: Dietitians in Saudi Arabia generally exhibited favorable attitudes toward food sustainability, with a gap identified in factual knowledge despite an overestimation of the perceived knowledge and practices pertaining to environmental sustainability, and these gaps did not consistently translate into sustainable food consumption behaviors. These findings highlight the importance of incorporating food sustainability, including environmentally related concepts, into nutrition education and professional development as a potential strategy to bridge the gap between knowledge and application in dietetic practices. Full article

Deep reinforcement learning (DRL) has become an important approach for manufacturing scheduling because it supports sequential decision-making under complex and changing production conditions. However, existing reviews primarily organize the literature by scheduling problem or learning method, providing less explicit support for tracing how manufacturing context, Markov Decision Process (MDP) formulation, scheduler architecture, and evaluation choices interact across heterogeneous studies. This study presents an evidence-based systematic literature review of DRL for manufacturing scheduling using a structured methodology for corpus construction, configuration-level coding, evidence traceability, study-quality assessment, and cross-study synthesis. The validated corpus comprises 52 primary studies and 54 independently coded DRL configurations. The evidence is synthesized across manufacturing scheduling characteristics, MDP design, DRL scheduler design, hybrid optimization, and empirical evaluation. The results show that scheduler design is context-dependent and architecturally diverse: manufacturing requirements are associated with differences in state, action, and reward formulation, while DRL schedulers combine different learning algorithms, representation architectures, control structures, and complementary optimization mechanisms. The evidence does not establish universal superiority for individual representations, algorithms, or hybrid architectures because reported outcomes remain strongly conditioned by problem formulation and experimental design. Evaluation evidence further highlights limited generalization, uneven statistical and component-level validation, and a continuing gap between benchmark or simulation studies and live industrial deployment. By linking study-, configuration-, and evidence-level information, this review provides a traceable basis for interpreting methodological relationships, identifying research gaps, and guiding the development and evaluation of DRL-based manufacturing scheduling systems. Full article

Camera-only BEV 3D object detectors are trained under highly imbalanced category distributions, and their matched object queries can exhibit directional semantic errors toward frequent classes. We investigate this behavior as a diagnostic problem: given fixed geometric predictions and fixed query–ground-truth assignments, how much class information remains accessible in frozen decoder features, which errors can be recovered, and where does recovery fail? We establish a scene-disjoint protocol in which recovery fitting and model selection use separate subsets of the official nuScenes training set, while all 150 validation scenes (6019 samples) remain final-only until all model and post-processing choices are fixed. Geometry-only Hungarian matching produces 158,253 fixed positive pairs on the full validation set. The frozen detector obtains a macro accuracy of 0.6136 on these pairs, while a lightweight factorized head trained on frozen features from decoder layer 4 reaches 0.6959 ± 0.0012 across three seeds. A linear probe achieves a macro accuracy of 0.8628 on the internal tuning split, whereas a shuffled-label control remains at chance (0.1000), indicating that substantial class information remains decodable from the frozen features. Tail-focused analysis further shows that recovered errors are more separable in frozen feature space than unrecovered errors across all 15 class-by-seed comparisons. However, recovery is not consistently observed across the controlled ResNet-18 and ResNet-50 configurations, and locked end-to-end evaluation decreases mAP from 0.2565 to 0.1620 and NDS from 0.3582 to 0.2796. These results support a geometry-controlled diagnosis of partial and class-dependent semantic recoverability, rather than improved localization, architecture-independent recovery, or deployable detection performance. Full article

This study analyzed 27 publications on material passports (MPs) in construction, published between 2019 and 2025. Literature-derived keywords were extracted using predefined criteria, normalized, classified into four analytical dimensions, and examined through co-occurrence networks generated with VOSviewer. Based on their prominence and associations, the authors propose a four-dimensional framework linking MP development objectives, information requirements, enabling technologies, and implementation challenges. This framework is an author-developed conceptual synthesis grounded in the literature; it is neither automatically generated by VOSviewer nor an empirically validated implementation model. Within the dataset, reuse emerges as the most prominent objective, associated with lifecycle management, traceability, information continuity, and decision support. BIM occupies a central position among enabling technologies, while interoperability represents a major implementation challenge. By integrating these findings, the framework explains how MP objectives guide information selection, how information requirements influence technology choices, and how implementation constraints provide feedback that may reshape other dimensions. Its contribution lies in transforming fragmented evidence into an interconnected, objective-driven analytical structure and providing a conceptual basis for minimum viable information implementation. Future pilot projects and comparative case studies are necessary to validate its engineering applicability. Full article