Search Results (121)

Crop diseases and pests have a significant impact on planting costs and crop yields and, in severe cases, can threaten food security and farmers’ incomes. Currently, most researchers employ various deep learning methods, such as the YOLO series algorithms and U-Net and its variants, for the detection of agricultural plant diseases. However, the existing algorithms suffer from insufficient interpretability and are limited to linear modeling, which can lead to issues such as trust crises in current technologies, restricted applications and difficulties in tracing and correcting errors. To address these issues, a dual-module Kolmogorov–Arnold Network (U-PKAN) is proposed for agricultural plant disease detection in this paper. A KAN encoder–decoder structure is adopted to construct the network. To ensure the network fully extracts features, two different modules, namely Patchembed-KAN (P-KAN) and Decoder-KAN (D-KAN), are designed. To enhance the network’s feature fusion capability, a KAN-based symmetrical structure for skip connections is designed. The proposed method places learnable activation functions on weights, enabling it to achieve higher accuracy with fewer parameters. Moreover, it can reveal the compositional structure and variable dependencies of synthetic datasets through symbolic formulas, thus exhibiting excellent interpretability. A field corn disease image dataset was collected and constructed. Additionally, the performance of the U-PKAN model was verified using the open plant disease dataset PlantDoc and a gear pitting dataset. To better understand the performance differences between different methods, U-PKAN was compared with U-KAN, U-Net, AttUNet, and U-Net++ models for performance benchmarking. IoU and the Dice coefficient were chosen as evaluation metrics. The experimental results demonstrate that the proposed method achieves faster convergence and higher segmentation accuracy. Overall, the proposed method demonstrates outstanding performance in aspects such as function approximation, global perception, interpretability and computational efficiency. Full article

Background/Objectives: Artificial intelligence (AI)-based chatbots present a viable approach to overcoming several challenges associated with conventional psychotherapy, such as high financial costs, limited access to mental health professionals, and geographical or logistical barriers. Thus, these chatbots are increasingly employed as complementary tools to traditional therapeutic practices in mental health care. Our aim was to develop and validate a scale to measure attitudes toward the use of AI-based chatbots for mental health support, i.e., the Artificial Intelligence in Mental Health Scale (AIMHS). Methods: A multidisciplinary panel of experts assessed the content validity. To confirm face validity, we carried out cognitive interviews and calculated the item-level face validity index. We applied factor analysis to verify the construct structure. We assessed measurement invariance across demographic subgroups. Concurrent validity was evaluated using three valid instruments. Reliability was tested through Cronbach’s alpha, Cohen’s kappa, and the intraclass correlation coefficient. Results: Factor analysis supported a two-factor five-item model. The two factors were technical and personal advantages, and explained 81.28% of the variance. The AIMHS demonstrated adequate concurrent validity, evidenced by statistically significant correlations with Artificial Intelligence Attitude Scale (r = 0.405, p-value < 0.001), Attitudes Towards Artificial Intelligence Scale (acceptance subscale; r = 0.401, p-value < 0.001, fear subscale; r = −0.151, p-value = 0.002), and Short Trust in Automation Scale (r = 0.450, p-value < 0.001). Configural, metric and scalar invariance were supported by our findings. Cronbach’s alpha was 0.798, and intraclass correlation coefficient was 0.938. Cohen’s kappa for the five items ranged from 0.760 to 0.848. Conclusions: The AIMHS is a five-item psychometrically sound and user-friendly instrument capturing two dimensions; technical and personal advantages. Future research should be undertaken to further evaluate the psychometric properties of the AIMHS across diverse populations and contexts. Full article

Second-order derivative information, including mixed curvature, is central to Newton and trust-region optimization, uncertainty quantification, and simulation-based design. Classical finite differences (FD) remain popular but require delicate step-size tuning and can suffer from cancelation and noise amplification. Complex-step differentiation offers machine-precision gradients without subtractive cancelation, yet many second-derivative complex-step formulas reintroduce differencing. Hyper-dual numbers provide an algebraically principled alternative: by lifting real code to a four-component commutative nilpotent algebra, one obtains exact first and mixed second derivatives from a single evaluation, without finite differencing. This article gives a consolidated theoretical and experimental foundation for hyper-dual numbers. We formalize the algebra, prove exact Taylor truncation at second order, derive coefficient–extraction formulas for gradients and Hessians, and state assumptions for exactness, including limitations at non-smooth points and the need to branch on real parts. We present implementation patterns and language skeletons (C++, Python 3.11.5, MATLAB R2023b), and we provide fair numerical comparisons with FD, complex-step, and AD baselines. Stability tests under additive noise and ill-conditioning, together with runtime and memory profiling, demonstrate that hyper-dual coefficients are robust and reproducible in floating-point arithmetic, particularly for black-box codes where Hessian information is needed but differencing is fragile. Full article

Data-driven marketing analytics has advanced targeting and optimization, yet its underlying infrastructure now functions as a complex sociotechnical system with overlooked ecological costs. This study conceptualizes programmatic advertising through a systems lens. It introduces the Ecological Delivery Paradox, a structural incongruity where environmentally friendly advertising messages are transmitted via energy-intensive delivery pipelines. Using an interpretivist–abductive design, we conducted 38 in-depth interviews with consumers and professionals, which were analyzed using reflexive thematic analysis in MAXQDA. Results show that awareness of hidden delivery costs emerges through a concretization threshold and crystallizes into metaphors such as “clean message, dirty conduit,” which trigger differentiated cognitive–affective pathways. These pathways shape trust trajectories across four profiles: cliff erosion, slow seep, suspended risk, and resilient cores. System-level moderators, including rationalization buffers, efficiency beliefs, and the visibility of low-data alternatives, determine outcomes. The findings extend marketing systems theory by reframing greenwashing as message–infrastructure misalignment and by integrating delivery congruence into advertising trust models. We propose a data-driven control architecture that aligns predictive analytics with ecological proportionality through mechanisms such as lightweight creatives, carbon-aware bidding coefficients, frequency–data quotas, and ad-level transparency labels. This systemic approach advances legitimacy, audience trust, and sustainability as joint objectives in programmatic advertising. Full article

Driven by government subsidies and advertising revenue, air taxis present an innovative solution to alleviate traffic congestion and are poised for growth. However, at their current stage of development, air taxi companies primarily operate short-distance routes within cities and rarely offer intercity services. Moreover, as a new mode of transportation, air taxis experience low levels of consumer trust at present. This study, grounded in the Hotelling model, examines differentiated decision-making scenarios between two competing air taxi service providers. It systematically analyzes how service expansion (specifically, the introduction of intercity services) and advertising strategies affect pricing, market share, and profits. Furthermore, it explores optimal decision-making patterns under external disturbances, providing theoretical support for service providers formulating operational strategies. We constructed a differentiated decision-making game model to simulate competition between Service Provider 1 (which does not offer intercity services but may advertise) and Service Provider 2 (which advertises but may choose whether to offer intercity services). By comparing game equilibrium outcomes under different decision combinations, we identify threshold conditions for key variables (e.g., additional price for intercity services and the advertising discount coefficient). The model is further expanded to incorporate external disturbance factors, allowing for analysis of how such environments influence the profitability of each decision pattern. Research has revealed that 1. offering intercity services can increase a provider’s optimal price and market share, but only if the “additional price for intercity services exceeds the threshold”; 2. both providers choosing advertising services is the optimal strategy, but if the advertising discount coefficient exceeds a reasonable range, it will intensify vicious competition. Therefore, it must be controlled within the optimal threshold to avoid adverse effects; 3. under external disturbance conditions, service providers prefer models that do not involve intercity services, and the “both parties advertise (NTX)” combination is more optimal. If intercity services are necessary, disturbance risks must be carefully assessed, or flexible cost and operational strategies should be implemented to hedge against negative impacts. Full article

Cost overruns remain one of the most persistent challenges in construction and infrastructure project management, often undermining efficiency, sustainability, and stakeholder trust. With the rise of digital transformation, artificial intelligence (AI) and machine learning (ML) provide new opportunities to enhance predictive decision-making and strengthen project control. This study introduces a digital project management framework that integrates the Pelican Optimization Algorithm (POA) with Light Gradient Boosting Machine (LGBM) to deliver reliable and interpretable cost overrun forecasting. The proposed POA-LightGBM model leverages metaheuristic-driven hyperparameter optimization to improve predictive performance and generalization. A comprehensive evaluation using multiple error metrics Coefficient of Determination (R²), Root Mean Squared Error (RMSE), Mean Absolute Error (MAE), and Mean Absolute Percentage Error (MAPE) demonstrates that POA-LGBM significantly outperformed baseline LGBM and alternative metaheuristic configurations, achieving an average R² of 0.9786. To support transparency in digital project environments, SHapley Additive exPlanations (SHAPs) were employed to identify dominant drivers of cost overruns, including actual project cost, energy consumption, schedule deviation, and material usage. By embedding AI-enabled predictive analytics into digital project management practices, this study contributes to advancing digital transformation in project delivery, offering actionable insights for cost control, risk management, and sustainable infrastructure development. Full article

Regional economic diversity and unevenly allocated space-based resources have created unprecedented difficulties for collaborative and innovative supply chain construction. This paper sets up a tripartite evolutionary model of the government, upstream companies, and downstream companies to explore dynamic processes of regional supply chain collaborative innovation with bounded rationality. Through incorporation of hierarchical space organizations and policy incentive differentiation mechanisms, the model discerns actors’ behavioral evolution and strategic adjustment in a geographically divided structure. Adopting evolutionary game theory and numerical simulation, this paper includes crucial parameters like the conversion efficiency of return conversion, information-sharing coefficient, mutual trust coefficient, and fiscal subsidy coefficient for examining policy and spatial heterogeneity effects on information collaborative innovations. The results reveal that fiscal incentives are the primary driving factor for collaborative evolution across local supply chains. Adaptive profit-sharing and subsidy intensities both stimulate upstream innovation investments and downstream cooperation adoption efficiently, stimulating a shift out of inefficient equilibrium states towards sustainable high-cooperation states. Furthermore, the restructuring of space accelerates hierarchical differentiation—core region companies are able to act like initiators and leaders for collaborative innovations, while periphery companies encounter participatory barriers in terms of elevated coordination costs and incentive shortages. In light of this, it is therefore crucial to have a “core-driven, periphery-subsidized” policy system for eliminating spatial gaps, stimulating cross-regional information exchange, and building systemic robustness. These findings contribute to enhancing the overall efficiency, stability, and innovation capacity of regional supply chain systems. They also provide a theoretical basis for policy decision making and industrial upgrading across regions of varying scales and environments. Full article

Objectives: Colorectal cancer (CRC) is the second-deadliest cancer globally, with an estimated 52,900 additional deaths expected in the United States by 2025. Early detection through colonoscopy significantly reduces CRC mortality by enabling the removal of pre-cancerous polyps. However, manual visual inspection of colonoscopy images is time-consuming, tedious, and prone to human error. This study aims to develop an automated and reliable polyp segmentation and classification method to improve CRC screening. Methods: We propose a novel deep learning architecture called µ-Net for accurate polyp segmentation in colonoscopy images. The model was trained and evaluated using the Kvasir-SEG dataset. To ensure transparency and reliability, we incorporated Explainable AI (XAI) techniques, including saliency maps and Grad-CAM, to highlight regions of interest and interpret the model’s decision-making process. Results: The µ-Net model achieved a Dice coefficient of 94.02%, outperforming other available segmentation models in accuracy, indicating its strong potential for clinical deployment. Integrating XAI provided meaningful visual explanations, enhancing trust in model predictions. Conclusions: The proposed µ-Net framework significantly improves the Precision and efficiency of automated polyp screening. Its ability to segment, classify, and interpret colonoscopy images enables early detection and supports clinical decision-making. This comprehensive approach offers a valuable tool for CRC prevention, ultimately contributing to better patient outcomes. Full article

Economic development and its sustainability are influenced not only by material, human, financial, and intellectual factors, but also by psychological factors. In particular, the levels of business expectations, trust, and confidence significantly affect the resilience of the economy, especially in crucial sectors such as industry and, more specifically, industrial production. Based on political, economic, social, and legal stability, businesses are likely to assess their opportunities more optimistically and realistically. This, in turn, enables them to look confidently toward the future and provides a foundation for investing in further development. Conversely, a decline in business expectations and confidence can slow socio-economic development, potentially leading to recession or depression. The purpose of the article is to identify the association between business confidence (Impact of the Business Confidence Indicator, IBCI) and the level of industrial production (Industrial Production Index, IPI), as a crucial aspect of ensuring sustainable economic development. A correlation–regression analysis conducted using Ukraine as a case study—a country candidate for EU accession—and statistical data from the State Statistics Service of Ukraine (SSSU) for the period from 1 February 2022 to 1 September 2024 demonstrated that there is a stable, positive, and strong relationship between IBCI and IPI levels (r = 0.7; D = 0.49). The constructed linear correlation model indicates that, with other factors held constant, a one-percentage-point increase in positive business expectations may lead to a 2.23-point rise in the industrial production activity of enterprises in Ukraine’s manufacturing sector. Furthermore, approximately 49.0% of the variation in industrial production levels is likely explained by changes in business expectations. Verification of the constructed regression equation and assessment of its parameters indicate that it is statistically reliable and consistent with real economic processes. Specifically, the Fisher coefficient (F = 5.30) exceeds the critical (tabular) value (Ft = 2.04), with Se = 0.45 and C_95% = 1.96; the causality test based on the Granger methodology revealed the presence of a causal relationship, indicating that the IBCI influences the IPI. The obtained statistical results for the applied models and tests are as follows: MDF (p < 0.05), KPSS (p > 0.10), Durbin–Watson ≈ 2.0, Breusch–Godfrey (p = 0.32), White (p = 0.41), ARCH (p = 0.27), and SER (p = 0.36). The constructed correlation–regression equation also allowed forecasting based on trend line modeling—how IPI levels will change depending on business confidence. According to the forecast, the IPI in Ukraine at the beginning of 2030 is expected to increase by 63.48 percentage points compared to the beginning of 2024, reaching 153.6%. Full article

This study analyzes how socio-demographic factors influence the sustainable consumption of organic vegetables in Romania. Following behavioral theories such as the Theory of Planned Behavior and Value-Belief-Norm models, the study explores how attitudes, trust, and socio-demographic characteristics shape individual purchasing decisions. The main objective was to analyze the frequency of consumption of organic vegetables, the reasons for action, the level of trust in organic certification and preferences regarding distribution channels in relation to socio-demographic variables. The research was based on a structured questionnaire applied to a sample of 533 respondents, selected from various regions of the country. The statistical analysis included descriptive and inferential methods, namely the Chi-square test for the association between variables, the Spearman coefficient for ordinal correlations and the Mann–Whitney U test for the comparison of independent groups. Results show that the frequency of organic vegetable consumption decreases as income and education rise, a pattern influenced by Romania’s market structure and cultural context, where higher-income consumers often prefer imported or internationally certified products, while middle-income groups sustain local purchases. Health is the main motivation for the purchase, while aspects related to environmental protection and support for local producers are mentioned secondarily. Price perception is significantly associated with age, income, and education, and the preference for purchasing channels especially depends on educational level. Trust in producers correlates with gender and income, while trust in certification labels shows no significant associations. These findings provide insight into how socio-demographic characteristics shape sustainable consumption behaviors and build a solid foundation for the development of the organic products market in Romania. Full article

Interest in consumer behavior regarding agri-food products is growing, particularly in the context of sustainable and local consumption. This study examines consumer perceptions of certified mountain products in Romania, with a specific focus on cow’s milk. A structured survey was conducted among 576 respondents from the Bucharest–Ilfov metropolitan area, representing the target population of regular food consumers. The data were analyzed using descriptive and comparative statistical methods in SPSS, including chi-square tests and contingency coefficients, to evaluate consumer awareness, attitudes, and their willingness to pay. The results reveal that although awareness of the “mountain product” label is high (88.9%), its direct influence on purchase decisions remains limited, with price, brand, and origin playing stronger roles. Nevertheless, 95% of respondents expressed willingness to pay a premium price, and over 70% associated mountain certification with health benefits and higher quality. These findings highlight both the potential and current limitations of certification as a market-based tool to support sustainable consumption and the economic resilience of mountain areas. The study contributes empirical evidence from a less explored national context. It offers insights for policymakers, producers, and retailers seeking to strengthen short food supply chains and consumer trust in certified labels. Full article

Background/Objectives: Although it is known that widespread online misinformation has negative consequences, there is no scale to measure susceptibility to online misinformation. Thus, our aim was to develop and validate a tool for measuring susceptibility to online misinformation: the Online Misinformation Susceptibility Scale (OMISS). Methods: A comprehensive literature review was conducted to generate a preliminary pool of items. Subsequently, a multidisciplinary panel of experts assessed the content validity of these items. To establish face validity, cognitive interviews were performed. Both exploratory and confirmatory factor analyses were performed to verify the underlying construct structure of the OMISS. We examined the concurrent validity of the OMISS by using a fake news detection scale, the Trust in Scientists Scale, a single-item trust in scientists scale, the Conspiracy Mentality Questionnaire, and a single-item conspiracy belief scale. Reliability was rigorously examined using multiple indices, including Cronbach’s alpha, McDonald’s Omega, Cohen’s kappa, and intraclass correlation coefficient. Results: The psychometric evaluation using both exploratory and confirmatory factor analyses supported a one-factor nine-item model for the OMISS. The OMISS demonstrated strong concurrent validity, evidenced by statistically significant correlations with the five scales mentioned above. Cronbach’s alpha and McDonald’s Omega were 0.920 and 0.921, respectively. The intraclass correlation coefficient for the OMISS was 0.994. Conclusions: Our comprehensive psychometric evaluations confirmed the OMISS as a valid tool for measuring online misinformation susceptibility. The OMISS holds promise as an effective tool for identifying susceptibility to misinformation and could support policymakers, health educators, healthcare professionals, and stakeholders in pinpointing high-risk groups. Full article

Background and Objectives: This study aims to evaluate the diagnostic potential of routinely available hematological parameters for acute myocardial infarction (AMI) by employing an Explainable Neural Network (ENN) model that combines high predictive accuracy with interpretability. Materials and Methods: A publicly available dataset comprising 981 individuals (477 AMI patients and 504 controls) was analyzed. A broad set of hematological features—including white blood cell subtypes, red cell indices, and platelet-based markers—was used to train an ENN model. Bootstrap resampling was applied to enhance model generalizability. The model’s performance was assessed using standard classification metrics such as accuracy, sensitivity, specificity, F1-score, and Matthews Correlation Coefficient (MCC). SHapley Additive exPlanations (SHAP) were employed to provide both global and individualized insights into feature contributions. Results: The study analyzed hematological and biochemical parameters of 981 individuals. The explainable neural network (ENN) model demonstrated excellent diagnostic performance, achieving an accuracy of 94.1%, balanced accuracy of 94.2%, F1-score of 93.9%, and MCC of 0.883. The AUC was 0.96, confirming strong discriminative ability. SHAP-based explainability analyses highlighted neutrophils (NEU), white blood cells (WBC), RDW-CV, basophils (BA), and lymphocytes (LY) as the most influential predictors. Individual- and class-level SHAP evaluations revealed that inflammatory and erythrocyte-related parameters played decisive roles in AMI classification, while distributional analyses showed narrower parameter ranges in healthy individuals and greater heterogeneity among patients. Conclusions: The findings suggest that cost-effective, non-invasive blood parameters can be effectively utilized within interpretable AI frameworks to enhance AMI diagnosis. The integration of ENN with SHAP provides a dual benefit of diagnostic power and transparent rationale, facilitating clinician trust and real-world applicability. This scalable, explainable model offers a clinically viable decision-support tool aligned with the principles of precision medicine and ethical AI. Full article

AI-powered health assistants offer promising opportunities to enhance health management among older adults. However, real-world uptake remains limited, not only due to individual hesitation, but also because of complex interactions among users, platforms, and public policies. This study investigates the dynamic behavioral mechanisms behind adoption in aging populations using a tripartite evolutionary game model. Based on replicator dynamics, the model simulates the strategic behaviors of older adults, platforms, and government. It identifies evolutionarily stable strategies, examines convergence patterns, and evaluates parameter sensitivity through a Jacobian matrix analysis. Results show that when adoption costs are high, platform trust is low, and government support is limited, the system tends to converge to a low-adoption equilibrium with poor service quality. In contrast, sufficient policy incentives, platform investment, and user trust can shift the system toward a high-adoption state. Trust coefficients and incentive intensity are especially influential in shaping system dynamics. This study proposes a novel framework for understanding the co-evolution of trust, service optimization, and institutional support. It emphasizes the importance of coordinated trust-building strategies and layered policy incentives to promote sustainable engagement with AI health technologies in aging societies. Full article

With rising environmental awareness, numerous firms are transitioning to green investment, such as low-carbon production. However, the consumer adoption of low-carbon products remains low due to transparency concerns. Many firms are leveraging blockchain to address information asymmetry in the supply chain, thereby building consumer confidence in low-carbon products. The purpose of this work is to provide decision support for business firms by analyzing the strategic choices regarding the manufacturer’s green investment and the e-retailer’s adoption of blockchain technology. Three strategy combinations are considered, including the baseline strategy combination without green investment and blockchain technology (NN), the strategy combination with only green investment (LN), and the strategy combination with both green investment and blockchain technology (LB). The optimal pricing and green level decisions are derived, and the conditions under which green investment and blockchain technology are beneficial to the supply chain members are examined. The findings suggest that the e-retailer can obtain the highest profit without adopting blockchain technology if it holds a substantial or extremely low market share, if the consumers’ low-carbon preference is at a low to medium level, or if the consumer green trust coefficient is high when the manufacturer implements the green investment strategy. When consumers exhibit a weak preference for low-carbon products, the strategy combination NN is optimal for the supply chain members. The strategy combination LB becomes optimal if the consumer green trust coefficient is near or below the moderate threshold, if the market share of a channel is neither extremely high nor low, or if consumers exhibit a strong preference for low-carbon products. Full article