This paper proposes a comprehensive infant undercount estimator designed to capture the full spectrum of infant undercounts in the population census, thereby replacing the currently widely used single-source undercount estimator, which is known to significantly underestimate the true scale of infant undercount. Employing an integrated methodology that combines mathematical modeling, multiple sampling techniques, and field surveys, the study addresses the construction and related issues of the comprehensive infant undercount estimator. Both theoretical and empirical findings demonstrate that the single-source undercount estimator fails to provide complete coverage of infant undercounts in the total population. Furthermore, the comprehensive infant undercount estimator must be constructed on the basis of homogeneous population strata; otherwise, it is susceptible to heterogeneity bias. Compared to single source, double source, and triple source undercount estimators, the comprehensive infant undercount estimator exhibits superior estimation accuracy, making it suitable for estimating infant undercounts. As a biased estimator, its precision should be assessed using its mean squared error estimator. This research makes an original contribution by introducing the comprehensive infant undercount estimator and establishing a systematic framework for estimating infant undercounts in population censuses. The comprehensive infant undercount estimator is expected to provide theoretical support and practical guidance for the National Bureau of Statistics in designing future infant undercount estimation programs, thereby enhancing their scientific rigor and operational feasibility.

Accurate forecasting of commodity prices helps policymakers respond promptly to external economic shocks and enhances macroeconomic resilience. This paper investigates the interdependencies within commodity markets and the transmission of risks across different assets, with a focus on price dynamics. Based on Bayesian inference principles, a prior heterogeneous graph is constructed incorporating three types of semantic relationships (causal, similarity, and correlation) to capture multidimensional influence pathways among commodities. Serving as the foundation for spatial graph convolution layers, this prior graph enables the model to effectively learn complex spatial dependencies among commodities. Leveraging an attention mechanism, self-attention value matrix is derived as posterior dynamic dependency matrices, representing the evolving interdependencies among commodities under price fluctuations, thereby helping identify potential market linkages and risk transmission paths, such as spillover effects from upstream industrial goods to downstream products. Ultimately, we propose a heterogeneous graph convolutional neural network model based on multi-head self-attention mechanisms for price fluctuation prediction. Using 50 actively traded commodity futures contracts from China's major trading markets as case studies, we conduct comparative experiments and ablation analyses based on preprocessed real-world trading data. Experimental results show that our proposed algorithm outperforms existing methods across multiple evaluation metrics, reflecting its effectiveness and practical value.

Data in fields such as finance, industry, medicine, and meteorology commonly exhibit characteristics including heteroscedasticity, skewness, and heavy-tailed distributions. The identification and testing of change-points in such data holds significant practical relevance and theoretical value. This study focuses on the joint location and scale models. Utilizing a modified information criterion (MIC), we propose a method for identifying change-point locations under the assumption of a skew-t-normal(StN) distribution. To evaluate the performance of this method, we conduct Monte Carlo simulations involving simultaneous shifts in both the location parameter and scale parameter, comparing the proposed method against the classical likelihood ratio test (LRT). Simulation studies demonstrate the superior performance of the proposed method over the LRT. To further validate the method, we applied it to air quality data recorded in Lanzhou, China, from September 1, 2023, to August 31, 2024. The two change-points identified through this analysis coincide with documented environmentally significant change-points, demonstrating the high accuracy of the proposed change-point detection and identification methodology.

To address the challenges of strong model uncertainty, insufficient generalization under complex operating conditions, and high operational cost in the denitrification dosing control of wastewater treatment plants, this paper proposes a mechanism-guided residual learning and flexible constraint optimization method. In this approach, a mechanistic model is employed to characterize the underlying dynamics of the dosing process, while LightGBM is utilized to learn and compensate for the residuals of the mechanistic model. Lag terms and temporal features are further incorporated to enhance the model's representational capacity. Building upon this hybrid framework, a prediction-based flexible dosage-saving strategy is introduced to achieve economic optimization of the dosing process. Experimental results demonstrate that the proposed hybrid model substantially outperforms the mechanistic model in predictive accuracy, reducing the Mean Absolute Percentage Error (MAPE) from 27.26% to 5.79%. Compared with purely data-driven models, it also exhibits superior out-of-distribution stability and generalization capability. Under Gap Temporal Split scenarios, the mechanism-guided model markedly mitigates performance degradation, while achieving a 3.54% reduction in chemical consumption under the constraint that effluent quality standards are satisfied. SHapley Additive exPlanations (SHAP) analysis further reveals that the incorporation of mechanistic structure enhances both the interpretability and operating-condition adaptability of the model. Overall, the proposed method achieves a favorable balance among predictive accuracy, robustness, and operational economy, offering an effective approach for mechanism-data fusion modeling and optimal control of complex wastewater treatment processes.

Air cargo serves as a crucial link in supporting the efficient operation of global supply chains, with its demand fluctuations driven by multiple complex factors. Based on the TEI@I methodology, this study applies complementary ensemble empirical mode decomposition to the monthly cargo data of ten major Chinese airports from 2015 to 2024. According to the stationarity and complexity of each subsequence, modeling and forecasting are conducted using seasonal autoregressive integrated moving average models, long short-term memory networks, and support vector regression models, with an event-driven adjustment module incorporated. Furthermore, this paper analyzes the correlation characteristics of cargo volumes within the northern, eastern, and southern airport clusters and compares differences in factor sensitivity and prediction error structures across different clusters. Empirical results demonstrate that the TEI@I prediction model constructed in this study performs well for most airports and evaluation indicators, providing reliable methodological support for multi-airport cargo volume forecasting. It also offers important reference value for strategic planning and resource allocation by local governments, airport operators, and air logistics enterprises.

China is one of the countries most severely affected by natural disasters, and emergency evacuation is a common and effective measure for disaster prevention and mitigation. In practice, emergency evacuations often require the rapid transfer of some vulnerable populations (e.g., the elderly, pregnant women, and the injured). However, post-disaster road networks may be damaged and congested, severely constraining evacuation efficiency. Electric Vertical Takeoff and Landing (eVTOL) aircraft feature vertical takeoff and landing, obstacle-crossing, and high-speed capabilities, enabling rapid transfer of vulnerable populations and complementing ground transportation to overcome road evacuation bottlenecks. Considering capacity differences and transfer service delays during ground-to-air mode transitions, this paper proposes a ground-air coordinated dynamic evacuation optimization model. In the model, the objective function is to minimize total system evacuation time by jointly deciding the ground-air splitting ratio and path traffic flow. First, the Cell Transmission Model (CTM) is used to characterize the dynamic evolution of ground-air evacuation traffic flows, and a transfer cell mechanism incorporating passenger capacity conversion and transfer service delays is designed. Second, a decomposition algorithm based on the Alternating Direction Method of Multipliers (ADMM) is proposed to solve our proposed model. Finally, the effectiveness of the model and algorithm is validated on the Sioux Falls network. The results indicate that: (1) ground-air collaborative evacuation significantly outperforms only using ground evacuation. The coordinated benefits show a concave growth trend with increasing evacuation demand and tend to stabilize. There exists a globally optimal splitting ratio (approximately 0.38 in the numerical example of this paper) to achieve dynamic resource matching; (2) ground-air transfer nodes are the core bottleneck restricting the evacuation efficiency of air corridors, where transfer waiting time is significantly longer than flight time and increases at a faster rate; (3) improving the passenger capacity of eVTOL and the service capacity of take-off and landing points can reduce the total evacuation time, but both exhibit diminishing marginal returns, and the optimal diversion ratio is more sensitive to changes in passenger capacity; (4) the takeoff and landing service capacity and transfer speed of ground-air transfer nodes determine the evacuation efficiency of the air corridor. The improvement of air transport capacity should be matched with transfer service capacity to improve the system's evacuation efficiency. This paper can provide a decision-making support for evacuation plans, route planning, and the allocation of eVTOLs and vehicles during an emergency evacuation.

Aiming at the problem of insufficient engineering accuracy caused by additive noise interference in weak signal detection of Duffing system, this paper models the deterministic Duffing system excited by signal and noise as a stochastic differential equation driven by additive Gauss noise. This paper analyzes the characteristics that the drift term of the system contains a cubic nonlinear term and satisfies the superlinear growth, and points out that it does not satisfy the linear growth condition required by the classical solution existence theory. To this end, this paper applies Mao's classical framework and generalized local Lipschitz condition system to the system, constructs a Lyapunov function that satisfies the Khasminskii condition, and strictly proves that the system has a global solution under any initial conditions. On this basis, the Lyapunov function with undetermined parameters is further constructed, and the explicit upper bound of the second-order moment of the solution is derived by combining It\^{o} formula and Bihari inequality. Then, the quantitative conclusion that the upper limit of the second-order moment Lyapunov exponent is not positive is obtained directly from the upper bound of the second-order growth, which shows that the system will not appear exponential runaway growth.

This paper proposes a cooperative control algorithm for multi omnidirectional mobile robot systems over directed topologies based on the Udwadia-Kalaba (U-K) approach, aiming to achieve formation tracking and obstacle avoidance control. First, within a leader-follower framework, the formation tracking task is formulated as equality constraints and further reformulated into a second-order differential form to match the acceleration-level constraint paradigm of the U-K method. Then, under the condition that the initial states lie within the safe set, obstacle avoidance inequality constraints are transformed into second-order constraints that can be embedded into the U-K framework by constructing control barrier functions. On this basis, the formation tracking and obstacle avoidance constraints are integrated to establish a weighted composite constraint set. By assigning different weights, safety-critical obstacle avoidance is approximately prioritized under finite weights, while a lexicographic solution with obstacle avoidance priority is obtained as the weight tends to infinity. Finally, an explicit control law is derived in the sense of weighted least squares based on the U-K method. Simulation results demonstrate that, in environments containing both static and moving obstacles, the proposed algorithm can restore formation tracking while satisfying safety distance constraints. Compared with a Lyapunov-based model predictive control method, the proposed algorithm exhibits superior real-time performance and control accuracy.

As the core driver of high-quality development in digital finance, the digital transformation of financial institutions requires in-depth research into its evolutionary mechanisms and implementation pathways. Based on the framework of evolutionary game theory, this paper constructs a three-party stochastic evolutionary game model involving local governments, financial institutions, and fintech companies. It systematically reveals the dynamic evolutionary mechanisms and strategic choice processes underlying the digital transformation of financial institutions, with a particular focus on analyzing the differential impacts of external environmental uncertainty on the transformation process. The study finds that: (1) the three parties exhibit nonlinear interactions during the evolutionary process, and their strategy choices are significantly interdependent; (2) local government policy interventions have a dual effect on the transformation process: moderate fiscal subsidies can significantly enhance the motivation for financial institutions to transform, but there is a critical threshold for policy effectiveness, and as market mechanisms improve, a policy substitution effect may occur; (3) Under deterministic evolutionary games, the optimal path for financial institutions’ digital transformation is characterized by “government guidance—market coordination—steady-state symbiosis”; however, this path is prone to deviation and distortion under stochastic evolutionary games, particularly during the market coordination phase. Under conditions of stochastic disturbances, the regulatory effectiveness of subsidy policies is significantly weakened, and fintech firms exhibit a stronger risk-averse tendency; The synergistic effects between financial institutions and fintech firms exhibit phased fluctuations, slowing the system’s convergence rate. Based on this, it is recommended that government departments establish a dynamically optimized policy toolkit, while financial institutions should refine their strategic plans for digital transformation. By deepening the dual-wheel synergy mechanism of “data empowerment + technology-driven innovation,” they can build a sustainable digital financial ecosystem.

Under the backdrop of intensifying pressure from global environmental governance and the accelerated advancement of the “dual carbon” goals, heavily polluting enterprises face higher requirements for environmental compliance, stricter emission supervision, and greater pressure to pursue green transformation. This makes them more inclined to engage in greenwashing in concealed and sophisticated ways, thereby gaining social recognition, brand premium, and policy inclinations through a false pro-environmental image. However, existing studies on greenwashing behavior identification mostly rely on single-modal data such as numerical or textual data, which not only lack the integrated application of multi-modal data but also generally overlook the impact of enterprises' top management affiliation relationships on the accuracy of greenwashing behavior identification. Therefore, this paper integrates numerical, textual, and affiliation network data to construct a greenwashing behavior identification model based on the Multi-modal Hierarchical Relational Graph Attention Network (MHRGAT). Empirical results based on 1,691 A-share listed heavily polluting enterprises in Shanghai and Shenzhen from 2014 to 2023 show that: (1) Compared with machine learning methods such as LR and RF, and deep learning models such as CNN and GAT, the MHRGAT model achieves the optimal identification performance. (2) The transmission efficiency of director affiliation information with different historical durations in the network varies, and selecting the director affiliation relationships of enterprises over the past 3 years is relatively appropriate for greenwashing behavior identification. (3) The MHRGAT model based on multi-modal fusion significantly outperforms models using single-modal or dual-modal data in identifying greenwashing behaviors of heavily polluting enterprises across multiple evaluation metrics, demonstrating that multi-modal data fusion enhances the overall performance of greenwashing detection. These research findings can provide regulatory authorities with a greenwashing behavior identification method based on multi-modal fusion and cross-modal attention, thereby guiding heavily polluting enterprises to actively fulfill their environmental responsibilities.

The energy transformation has promoted the development of the lithium-ion battery industry chain. There are complicated international trade relations among the participating economies in the industrial chain. Unexpected events such as geopolitical games make trade or production risks exist in all links of the industrial chain, and the spatio-temporal transmission of risks will pose a threat to the international trade resilience of the entire industrial chain. In the face of risks, positive or negative feedback from other participating economies will affect the international trade resilience of the industrial chain. However, in the current research on international trade in the lithium-ion battery industry chain, there is a lack of studies focusing on the impact of economies' feedback on trade resilience, and most of them do not take into account the production relations among products. Therefore, this paper constructs a “production-trade” multi-layer network model. Through scenario setting and simulation, the international trade resilience of the lithium-ion battery industry chain combined with the positive and negative feedback of the economy is measured, and the following conclusions are obtained: (1) China holds an important position in multiple links of the international trade industrial chain of lithium-ion batteries. Chile and Australia play important positions in the upstream trade. (2) Risk transmission varies in different links of the international trade industrial chain of lithium-ion batteries. Especially lithium carbonate, its risk impact can carry out a long spatio-temporal conduction on a global scale. (3) Positive feedback helps to improve the international trade resilience of the industrial chain, and vice versa. Positive and negative feedback has a more significant impact on the resilience of middle and upstream products. The research results can provide a reference for improving the toughness of international trade of lithium-ion battery industry chain. The research framework proposed in this article can be extended to other industrial chain studies.

This study focuses on two fundamental challenges in the real scenario of financial fraud detection: the limited generalization ability of traditional supervised models under highly imbalanced data, and the high costs of fraud identification. Using annual report data of Chinese A-share listed firms from 2012 to 2022, we construct a financial indicator system consisting of 31 indicators covering thirteen types of fraud schemes, guided by the extended Fraud Triangle Theory, regulatory concerns, and audit practice. Methodologically, this study proposes a generative financial fraud detection framework based on a Convolutional Variational Autoencoder (CNN-VAE). By learning the latent distribution of normal firms’ financial characteristics, the framework identifies anomalous samples and adapts well to scarce fraud samples and imbalanced class distributions. A retrieval-augmented anomaly scoring method is further introduced to characterize potential fraud risk, together with a threshold-optimization mechanism to maximize economic gains from detection. Empirical results show that the proposed method outperforms mainstream machine-learning and deep-learning models across most evaluation metrics, even when benchmark models are enhanced with resampling strategies. Cost analysis further indicates that threshold optimization substantially improves the economic value of fraud detection in the real scenario, enabling most models to achieve positive net benefits.

Exchange rates are characterized by nonlinearity, non-stationarity, and complex fluctuations. Existing methods often suffer from limited prediction accuracy due to single modeling perspectives and inadequate utilization of unstructured data. This paper proposes a combined exchange rate forecasting approach that integrates LSTM and GNN for spatiotemporal feature fusion, enhanced with sentiment analysis based on large language models. First, a pre-trained large model is employed to perform sentiment analysis on unstructured data, quantifying market sentiment scores. These scores are then integrated with historical exchange rate data and fed into a Long Short-Term Memory (LSTM) network to extract temporal dependency features. Second, based on economic theory, relevant macroeconomic indicators are selected to construct a multidimensional graph structure, and a Graph Neural Network (GNN) is used to model the spatial dependencies among variables. Finally, an optimal weighted combination method is applied to integrate the individual predictions, yielding the final exchange rate forecast. To validate the effectiveness of the proposed combined forecasting model, an empirical prediction analysis is conducted on the daily USD/RMB exchange rate from January 2018 to February 2025. The results demonstrate that the proposed method is suitable for forecasting nonlinearly fluctuating exchange rates and achieves higher accuracy compared to existing approaches.

Fractional factorial designs are widely used in industry, agriculture, biopharmaceuticals, high technology and other fields due to their high efficiency. Assessing the optimality of fractional factorial designs has always been an important problem in the area of experimental design. Based on absolute distance and incorporating level permutations of factors, this paper defines the average absolute-moment as a criterion for measuring and screening good designs. Analytical relationships between average absolute moment and generalized word length pattern, moments, and orthogonal vectors for three-level designs are established. Using the wordlength enumerator, a fast method for computing average absolute moment of three-level designs is provided, and its lower bound is derived. This lower bound can serve as a benchmark for evaluating the goodness of designs in various dimensions. Finally, numerical examples are presented to illustrate the theoretical results.

As intelligent transformation reshapes the global manufacturing competitive landscape, how traditional manufacturing firms upgrade to develop New Quality Productivity (NQP) has become a growing research focus. Grounded in a two-factor productivity decomposition perspective, this study develops a partial equilibrium model to examine the impact of intelligent manufacturing on firms' NQP and its underlying mechanisms. We then apply a difference-in-differences design to Chinese A-share listed manufacturing firms. The results show that intelligent manufacturing significantly promotes firms' NQP. Mechanism evidence indicates that intelligent manufacturing reshapes the production function and drives the transition toward NQP through three channels, namely robotics innovation, green innovation, and labor structure optimization, with additional gains arising from their complementarities. Heterogeneity analyses further show that the productivity upgrading effect is more pronounced for technology-intensive firms, capital-intensive firms, and firms in the growth stage. These findings provide theoretical and empirical support for the micro-level pathways through which intelligent manufacturing fosters NQP and inform policies for high-quality development in China's manufacturing sector.

In the era of the digital economy, digital transformation, as an emerging development model for microeconomic entities, has become a new driving force for the transformation, upgrading and high-quality development of traditional brick-and-mortar enterprises. From the perspective of accounting conservatism, this paper takes Shanghai and Shenzhen A-share listed companies from 2007 to 2021 as research samples and constructs a two-way fixed effects model for empirical testing. The results show that corporate digital transformation can significantly improve accounting conservatism, and the digital transformation of peer enterprises has a spillover effect on the improvement of accounting conservatism. In particular, the application of underlying digital technologies exerts a more prominent promoting effect on accounting conservatism. Mechanism tests indicate that corporate digital transformation improves accounting conservatism through multiple channels: increasing corporate information transparency, enhancing internal control quality, reducing real earnings management behaviors, and lowering operational uncertainty. Heterogeneity analysis shows that the positive impact of digital transformation on accounting conservatism is more pronounced in enterprises with higher environmental uncertainty, enterprises in non-digital industries, and large-scale enterprises; meanwhile, the higher the industry information transparency, the more significant the industry spillover effect of digital transformation. This study enriches the literature on the economic consequences of corporate digital transformation and the influencing factors of accounting conservatism, and provides important reference and enlightenment for realizing coordinated industrial development against the background of digital transformation and helping traditional physical enterprises better capture digital dividends.

Considering the suddenness and uncertainty inherent in vaccine safety crises, this study employs a random stopping model to characterize the stochastic nature of crisis occurrences. This mechanism is incorporated into a vaccine supply chain system to construct a two-echelon differential game model comprising a vaccine manufacturer and a vaccination institution. Based on continuous dynamic programming theory, optimal strategies for information disclosure, promotion, and freshness investment are derived under four decision-making modes: vaccination institution introduction, manufacturer introduction, vaccination institution cost sharing, and manufacturer cost sharing. Furthermore, the impacts of key parameters on vaccine freshness, reputation, and profits are analyzed. The results indicate that:(1) Crisis shocks exhibit significant “pre-crisis inhibition” and “post-crisis forcing”effects. Specifically, high risk expectations dampen supply chain members’ ex-ante investment incentives, whereas high-intensity actual damage serves as a necessary condition for triggering the reverse recovery of reputation after a crisis. (2) The cost-sharing mechanism within a reasonable interval can effectively improve supply chain coordination performance, Notably, cost sharing by the downstream vaccination institution is more conducive to enhancing the production information transparency of the upstream manufacturer, thereby facilitating vaccine reputation recovery. (3) After the crisis, differences in the external introduction agents compel supply chain members to dynamically adjust their disclosure and promotion strategies according to specific crisis scenarios, thereby influencing vaccine trust levels and reputation recovery trajectories.

Citizen scientific literacy is an important indicator reflecting a nation's stock of scientific knowledge, scientific thinking ability, and scientific ethos. Against the backdrop of the deepening innovation-driven development strategy, improvements in citizen scientific literacy not only enhance public participation in innovation activities but also contribute to a more favorable regional innovation environment. However, existing studies have predominantly focused on the roles of R\&D personnel and higher education in shaping innovation outcomes, while systematic quantitative analyses of citizen scientific literacy remain relatively scarce. Based on panel data from 31 Chinese provinces, this study systematically examines the impact of citizen scientific literacy on regional innovation output, the results indicate that: (1) citizen scientific literacy has a significantly positive effect on regional innovation output, and mediation analysis further reveals that this effect is transmitted not only through direct channels but also through three indirect pathways—talent supply, innovation investment flows, and consumption demand—among which the mediating roles of consumption expansion and increased science and technology investment are particularly prominent; (2) regional heterogeneity analysis shows that the marginal effect of citizen scientific literacy on innovation output is significantly stronger in regions with higher concentrations of innovation factors, such as the central and eastern regions; (3) threshold effect analysis identifies a “double-threshold” relationship between citizen scientific literacy and regional innovation output, whereby the marginal effects exhibit a pattern of “increase-rapid increase-moderate increase” across low, medium, and high stages, with the strongest marginal impact occurring at the medium level of scientific literacy. Based on these findings, this study suggests that policies aimed at enhancing citizen scientific literacy should be tailored to regional development conditions and stage-specific characteristics, and should strengthen its role in supporting regional innovation capacity through talent formation, science and technology investment, and demand-induced innovation, thereby promoting the coordinated advancement of scientific literacy and innovation output.

In the context of the deep integration of digital technology and industrial chains, how can the digital transformation of enterprises enhance the production efficiency of upstream and downstream enterprises through the supply chain network? Especially, as the core "pioneer", can the digital practices of the chain leader drive the improvement of production efficiency among enterprises in the supply chain? This paper, based on the supply chain relationship data of A-share listed companies from 2009 to 2022, obtained the "enterprise - customer/supplier - year" samples through matching, and empirically analyzed the impact of digital transformation on the total factor productivity of enterprises in the supply chain and its transmission path. The research findings are as follows: First, digital transformation significantly improves the production efficiency of enterprises in the supply chain, and the chain leader plays a crucial reinforcing role in this process, exerting a "pioneer leading" effect. Second, the mechanism test shows that digital transformation mainly enhances the overall resilience of the industrial chain and supply chain by strengthening the structural and operational resilience of the supply chain, thereby promoting the improvement of total factor productivity of enterprises in the supply chain. Third, the heterogeneity test results further reveal that when enterprises are non-state-owned, high-tech industry enterprises, or when the concentration of the supply chain they are in is relatively high, the impact of digital transformation on the total factor productivity of supply chain enterprises is more significant. This study not only provides an explanation from the supply chain perspective for the spillover effect of digital transformation but also offers policy implications for the chain leader to lead the coordinated upgrade of the industrial chain.

Vision foundation models are difficult to apply to medical diagnosis directly and effectively due to the domain gap between medical images and pre-training natural images, and the scarcity of high-quality annotations. Model merging can integrate complementary knowledge from multiple foundation models, enhance model representation capability, and provide a new technical pathway for medical diagnosis. However, high-quality vision foundation model merging involves coupled optimization of discrete and continuous variables, such as source model layer selection and merging weight assignment, leading to an enormous search space; Meanwhile, performance evaluation of candidate merged architectures is expensive. To address the challenges in merging optimization, this paper formulates the search for the optimal fusion model as a mixed-integer programming problem. Leveraging the semantic correlation between network layers and CT diagnosis, we construct a network structural constraint of “shallow layers for characteristics—deep layers for diagnosis”, and design a layer-order consistency constraint between the merged and source models to preserve the conceptual evolution relationships implied in the hierarchical order of source models. To solve this optimization problem efficiently, we propose a simulated annealing algorithm with progressive perturbation mechanism and one-step gradient descent performance evaluation. By adaptively adjusting the perturbation magnitude to balance global exploration and local exploitation, and using single-step gradient descent to rapidly optimize merging weights and estimate candidate architecture performance, the algorithm improves both optimization efficiency and quality. This paper is the first to apply vision foundation model merging to pulmonary nodule CT diagnosis. Experimental results on multi-center datasets demonstrate that the proposed method outperforms four fine-tuning methods for single model training, four model merging rules, and two evolutionary algorithms in optimization performance, validating its effectiveness and superiority in the pulmonary nodule classification tasks.

Driven by “dual carbon” goals and the rapid growth in computing demand, the energy consumption and carbon emissions of data centers have become increasingly prominent concerns, making deep integration with renewable energy a key pathway for green and low-carbon transition. This paper addresses the deferability of computing tasks and the intermittency of solar and wind resources. A queuing-theory-based model for deferred scheduling of computing tasks is developed and embedded within a planning and operational optimization framework for wind-solar-storage systems, yielding an integrated co-optimization model for computing-power and electricity systems. This model achieves simultaneous optimization of micro-level computing task scheduling and macro-level wind-solar-storage planning. The resulting mixed-integer programming problem is solved using the Benders decomposition algorithm. Numerical results across three scenarios demonstrate that incorporating computing task scheduling reduces energy storage capacity requirements by 15.6%~54.6%, indicating a significant substitution effect of flexible computing loads on physical storage. Compared with single-resource scenarios, the wind-solar complementary scenario reduces total system cost by 54.1%~65.7% without scheduling; introducing computing task scheduling reduces total costs by $15%~50.5%$ across all three scenarios. Sensitivity analysis reveals that computing task scheduling substantially alters the sensitivity structure of system costs, with the wind-dominated scenario exhibiting the most pronounced improvement in the temporal matching between renewable generation and load. The proposed integrated planning and operation optimization model provides a theoretical framework and decision-support tool for the low-carbon transition of data centers.

To address the issue of insuffcient last-mile fulfillment capability for high-frequency sales services on e-commerce platforms, this study constructs a relaybased delivery crowdsourcing model that accounts for different arrival sequences of two couriers, and analyzes the sequence of“the first-arrived” and“the later-arrived” for couriers impact on the delivering effciency and profits of the e-commerce system. The findings show that: The strategy of relay-based delivering always achieves longer distances and higher service levels compared with the strategy of single-courier“pointto-point” . For two couriers with different cost structures, “the first-arrived” player wins the more tasks and profits in relay-based crowdsourcing scenario, which leads both couriers to motivate to actively compete for orders. From the perspective of the platform and the supply chain system, however, it is more beneficial if the lowcost (advantaged) courier arrives first and dominates the relay-based delivery, as this scenario yields a larger market sharing, higher sales, and greater system profits. To encourage the participation enthusiasm of disadvantaged(high-cost) couriers in relay delivery, the platform can adopt two mechanisms: An order assignment with revenue-sharing mechanism and commission rate tilt mechanism, Both mechanisms can effectively extend the system’ s delivery distance, expand the customer market coverage, and increase the profitability of the delivery system. Finally, numerical examples are used to verify the accuracy of the conclusions.

Traffic cooperation specifically refers to the formal establishment of a traffic transaction relationship between a social e-commerce platform (the traffic supplier) and a traditional e-commerce platform (the traffic demander), with the aim of achieving a win-win outcome for both participating parties. In a supply chain system consisting of three key entities—a social e-commerce platform (SEP), a traditional e-commerce platform (TEP), and a supplier— considering both the marketplace mode and the reselling mode of SEP, this paper systematically explores various traffic cooperation strategies between the two platforms under different operation modes and comprehensively analyzes their corresponding impacts on consumer surplus and overall social welfare. The main research findings show that when the product market outlook is moderate, the two platforms cooperate on traffic only in the reselling mode if the commission rate charged by SEP to the supplier is high, while a low commission rate leads to traffic cooperation only in the marketplace mode. When the market outlook is favorable, they cooperate in both operation modes; when unfavorable, they cooperate in neither operation mode. Moreover, when both SEP and TEP are willing to conduct traffic cooperation in both operation modes, SEP tends to cooperate in the marketplace mode, while TEP may show a clear preference for the reselling mode. In addition, platform interconnection based on traffic cooperation can effectively achieve a triple-win situation simultaneously among platforms, consumers, and the whole society under certain specific conditions.

Promoting the opening and sharing of government public data has become a strategic initiative to release the dividends of data factors and accelerate the cultivation of innovation momentum. Enhancing urban innovation resilience is an indispensable key link in accelerating the construction of resilient cities and promoting high-quality urban development. Therefore, systematically investigating the relationship between public data openness and urban innovation resilience holds significant value. This paper takes the successive launch of public data open platforms by Chinese local governments as a quasi-natural experiment and uses panel data of 281 Chinese cities from 2009 to 2023 as the research sample. Employing a time-varying difference-in-differences model, it systematically examines the driving effect of public data openness on urban innovation resilience. The results show that public data openness significantly improves urban innovation resilience, and this conclusion remains robust after endogeneity analysis and a series of robustness checks. Mechanism analysis reveals that the agglomeration of innovative talent and the increase in entrepreneurial activity serve as important bridging mechanisms through which public data openness drives urban innovation resilience, representing effective pathways for enhancing urban innovation resilience. Moreover, there exists a significant positive synergy between the agglomeration of innovative talent and entrepreneurial activity, and their interaction further strengthens urban innovation resilience. Heterogeneity analysis based on urban location shows that the driving effect of government public data openness on urban innovation resilience is stronger in eastern and coastal cities. Analysis based on city size indicates that public data openness has a more prominent enhancing effect on the innovation resilience of large cities. Analysis based on urban resource endowments reveals that the policy effect of public data openness is more pronounced in non-resource-based cities. Furthermore, analysis based on differences in urban digital infrastructure and industrial structure rationalization shows that the policy effect of public data openness is stronger in cities with more advanced digital infrastructure and higher levels of industrial structure rationalization. Therefore, local governments should fully recognize the innovation-driven effect of public data, accelerate the construction of data sharing platforms, formulate supporting policies related to entrepreneurship and talent agglomeration, implement stratified and categorized support systems, and actively leverage their own advantages to adopt a context-specific, step-by-step approach in formulating differentiated innovation development strategies.

This paper focuses on designing performance-based incentive contracts for China's contractual REITs, considering the risk aversion of both sponsors and managers as well as peer effects among managers. We develop a multi-agent principal-agent model where managers care about both absolute performance and returns relative to peers, and introduce reward and penalty coefficients to characterize symmetric and asymmetric relative-performance evaluation. On this basis, we derive the optimal incentive contract under different combinations of participation and incentive-compatibility constraints, and explore how the intensity and symmetry of peer comparison shape the equilibrium wage structure, managers' effort decisions and sponsors' expected payoffs. We further embed the model into a PPO reinforcement learning framework to capture the dynamic adjustment of incentive coefficients and wage levels. The results show that symmetric reward and penalty mechanisms effectively drive high effort and balance surplus allocation when managers' reservation utilities are low and market competition is intense. As managers' outside options improve, the optimal contract tends to be more punitive and rely more on fixed pay, which weakens incentive efficiency. PPO simulations also verify that excessive punishment leads to inefficient high-wage and low-effort outcomes, while moderately symmetric peer-comparison schemes improve effort incentives and meet fairness requirements. These findings offer practical implications for selecting relative-performance mechanisms and optimizing incentive and regulatory designs for REITs.

This study considers a Data Service Supply Chain “DSSC” consisting of a data provider, a data service provider, and users. Based on the unique characteristics of data such as infinite replicability and copyability and users' demand for data timeliness, we constructed two profit decision game models: Pricing Based on Data Timeliness “PB-DT” and Pricing Based on Usage Frequency “PB-UF”. By constructing profit decision game models based on two data asset pricing models, this study investigates the technology investment decisions of DSSC members, data asset pricing methods, and data asset and data service pricing decisions. Firstly, the study found that under both pricing models, the decisions of DSSC members do not affect the sales volume of data services. Furthermore, the sales volume of data services under the PB-DT pricing model is greater than that under the PB-UF pricing model. However, this does not necessarily mean that the profits of DSSC members and user utility are greater under the PB-DT pricing model than under the PB-UF pricing model. Secondly, the study found that the data supplier's choice of pricing method for data assets is solely related to the data service provider's valuation of the data's timeliness. For example, when the data service provider's valuation of the data asset is low (or high), the data supplier chooses the PB-UF (or PB-DT) pricing method. Finally, the study found that the data pricing method actually chosen by the data supplier may be “consistent” with or “contradictory” to the data pricing method expected by the data service provider and users. Especially, when the DSSC members have consistent preferences regarding the data asset pricing method, the synergistic effects among DSSC members are maximized, and the DSSC performance can reach a Pareto optimal state.

This paper investigates the design of a probabilistic set-membership filtering algorithm for nonlinear complex network systems under hybrid attack environments. Specifically, a distributed nonlinear estimation model is proposed to address measurement channels simultaneously subjected to false data injection (FDI) and denial-of-service (DoS) attacks. The measurement modeling mechanisms are analyzed under three scenarios: no attack, single attack, and hybrid attack. To handle the nonlinearities in the network dynamics, measurement equations, and FDI processes, the Taylor expansion technique is employed. Sufficient conditions for the existence of the probabilistic set-membership filter are derived, and a convex optimization problem is formulated. A recursive algorithm for calculating the filter gain matrix is then developed, ensuring that the estimation error remains constrained within a desired ellipsoidal region at a prescribed confidence level. Finally, numerical simulations are conducted to verify the effectiveness of the proposed filtering algorithm.

In the pellet shaft furnace process, stable control of the outlet temperature in the vertical roller mill combustion chamber is crucial to ensuring product quality. However, in actual production, the outlet temperature still relies on manual valve adjustment, making it difficult to maintain within the specified range. The inability to achieve automatic regulation arises primarily from two factors: first, fluctuations in gas pressure cause strong nonlinearity between the valve opening and gas flow rate; second, the flow sensor exhibits measurement saturation, leading to partial loss of state information. To address these issues, this paper divides the gas pressure variation range into several intervals, within which the relationship between valve opening and flow rate is assumed constant. Subsequently, separate models are established for the valve opening-flow system and the flow-outlet temperature system. An empirical-measure identification algorithm based on a measurement indicator vector and a regression identification algorithm based on adaptive gradient are proposed to achieve consistent parameter estimation. Using the identified models, a composite control algorithm with an accumulator structure is designed, under the assumption of a fixed valve adjustment increment, to stabilize the outlet temperature within the desired range. The experimental verification results show that the algorithm proposed in this paper can achieve parameter consistency, and the composite control algorithm outperforms the commonly used feedforward control algorithm on multiple statistical indicators.

Under the current context of the rapid development of agricultural product cold chain logistics and the in-depth advancement of the “dual carbon” goals, the agricultural supermarket direct supply system is confronted with multiple challenges such as the difficulty of multi-temperature layer co-distribution, the limited range of electric vehicles, and the uneven cost-sharing among multiple parties. Traditional logistics models often optimize from a single dimension, making it difficult to systematically integrate resources, achieve green collaboration, and save costs. Therefore, this paper proposes an integrated operation model of “multi-temperature co-distribution - electric vehicle routing - collaborative distribution” for agricultural supermarket direct supply. It aims to integrate multi-enterprise orders, vehicle and charging facility resources through a digital platform, build a dynamic collaborative alliance, and achieve systematic optimization of cold chain logistics. Firstly, an electric vehicle routing optimization model with the objective of minimizing total operating costs is constructed, comprehensively considering constraints such as multi-temperature layer loading, battery power, and charging strategies. Secondly, an adaptive large neighborhood search algorithm (ALNS) integrating spatio-temporal clustering is designed to efficiently solve the complex routing optimization problem, and the Shapley value method is adopted to fairly allocate the total cost within the collaborative alliance to ensure its stability and the enthusiasm of participating enterprises. Finally, the effectiveness of the model and algorithm is verified through numerical experiments and case studies. The results show that compared with the traditional collaborative distribution and multi-temperature co-distribution electric vehicle distribution models, the proposed model can achieve total cost savings of 15.29 % and 24.14 % respectively, and the average cost of each enterprise is reduced by 15.39 % and 24.19 % respectively.

This paper proposes a self-triggered distributed model predictive control algorithm to address the issue of high computational and communication resource consumption in existing distributed model predictive control algorithms for nonlinear multi-agent systems. At each triggering instant, each agent solves a local optimization problem based on its own local system state and the predicted states of its neighboring agents to determine its next triggering instant and broadcasts its predicted state trajectory to its neighbors. The algorithm significantly reduces communication load while maintaining system robustness and control performance. A decentralized nonlinear disturbance observer and a spatial decomposition technique are introduced to estimate and compensate for the matched disturbances in the system. On this basis, a robust contraction constraint is introduced in each local optimization problem to counteract the effect of residual disturbances. The recursive feasibility of the proposed algorithm and the closed-loop stability of the multi-agent systems at the triggering instants are proved. Numerical simulations verify the correctness of the theoretical results.

The snowball structured product is one of the most prominent products in China's over-the-counter (OTC) derivatives market. Its highly path-dependent structure presents significant challenges for pricing and risk management. This paper introduces the Willow Tree Method, a computational framework for pricing and calculating the Greeks for snowball products. This method is effective with various stochastic models, including geometric Brownian motion, jump-diffusion models, and stochastic volatility models. Compared to the traditional method, the willow tree method significantly improves computational efficiency and reduces costs while maintaining accuracy. Furthermore, this paper quantitatively assesses the instantaneous selling pressure of hedging positions under extreme scenarios and its impact on market liquidity, and demonstrates the framework's adaptability to other complex snowball structures. This expands the application prospects of the Willow Tree Method in the field of complex derivative pricing. It not only holds significant practical implications for financial institutions in managing risk and optimizing strategies within complex market environments but also provides quantitative references for regulatory authorities to prevent procyclical systemic risks induced by structured products.

The Panel Smooth Transition Regression (PSTR) model is widely applied in economics and finance to capture coefficient dynamics across regimes, but two limitations constrain its applicability: parameter estimates are sensitive to initial values and lack robustness, and the single-transition-variable specification cannot characterize the synergistic nonlinear influence of multiple explanatory variables on coefficient dynamics. To address these limitations, this paper proposes the Panel Multivariate Smooth Transition Regression (PMSTR) model, which replaces the polynomial expansion of a single transition variable with a multivariate logistic function. This specification simultaneously identifies the independent and synergistic moderating effects of multiple variables on regression coefficients and quantifies their relative moderating strengths. Parameter estimation employs a Bayesian inference framework combined with Hamiltonian Monte Carlo (HMC) sampling, in which prior regularization resolves the saturated solutions arising from the boundedness of the logistic function, while gradient information guides exploration of the high-dimensional non-convex parameter space. Monte Carlo simulations across four typical coefficient-dynamic scenarios---constant, discontinuous, linear, and nonlinear---demonstrate that PMSTR stably recovers coefficient trajectories and the relative influence strengths of variables; under a logistic data-generating process with one hundred independent replications, the model achieves strict parameter recovery with 89 % and 95 % posterior density interval coverage rates both meeting or exceeding nominal levels, whereas PSTR estimates exhibit substantial dependence on initial values. An empirical application following the analytical framework of Gonz\'{a}lez et al. examines investment--cash flow sensitivity for U.S. and Chinese listed firms. The results reveal that firm value is the dominant moderating variable in both samples, yet the synergistic moderation pattern between cash flow and firm value differs structurally across the two markets, underscoring the framework's explanatory capacity for complex economic and financial dynamics.

This paper constructs a duopoly supply chain model to examine the governance effect of digital technology on manufacturer’s greenwashing behavior under direct and resale channels, and analyze the impact of different sales channel choices on green innovation and corporate profits. The results show that the direct sales model provides the stronger incentive for green innovation in the absence of digital governance. After introducing digital technology, the channel advantage depends on the governance cost. The direct sales mode is more conductive to improve green levels under low governance cost, while the resale mode with cost-sharing mechanism becomes more advantageous under higher governance costs. Furthermore, when digital governance cost lies in some range, the resale mode can simultaneously enhance product green levels and green manufacturer’s profit. When the cost exceeds some critical threshold, green manufacturer’s profit falls below the no-governance scenario, while greenwashing manufacturer’s profit rises instead, leading to governance failure. Numerical simulations further demonstrate that as the proportion of prosocial consumers increases, the disciplining effect of digital governance on greenwashing behavior diminishes marginally, and excessively high governance costs may trigger an adverse selection where green manufacturer is harmed while greenwashing manufacturer benefits. The conclusions of this paper provide references for manufacturer’s sales channel selection and greenwashing governance decisions, and offer theoretical guidance for policymakers in designing cost subsidy mechanisms and differentiated regulatory strategies.

This paper employs a Stackelberg game model to investigate the effects of three government environmental policies-command-and-control, command-and-control with subsidies, and tax-and-subsidy policies-on enterprises’ capacity upgrade and emission reduction decisions. The results show that command-and-control policies directly constrain enterprises' emissions and indirectly encourage capacity upgrades; however, when the government sets a high target capacity upgrade ratio, such policies are less effective in promoting capacity upgrades. When the government imposes strict emission limits and high emission reduction requirements, only the command-and-control with subsidies policy can achieve the expected targets. Tax-and-subsidy policies can directly encourage capacity upgrades and indirectly reduce emissions; however, when the government budget is limited or when the emission reduction and capacity upgrade targets are relatively low, tax-and-subsidy policies are not cost-efficient, and command-and-control policies incur the lowest cost. These findings provide a policy selection basis for governments to promote enterprise capacity upgrading and pollution emission control.

Many practical problems in economics, management, engineering control, and artificial intelligence can be formulated directly or indirectly as unconstrained optimization problems, and the conjugate gradient method is one of the core algorithms for solving such large-scale problems. In this paper, we propose an improved Dai-Liao conjugate gradient method by modifying and truncating the Dai-Liao conjugate parameters and introducing a restart step into the search direction. Without relying on any line search conditions, the search direction generated by this method can satisfy the sufficient descent condition. Under standard assumptions, the strong convergence of the proposed method is proved. Numerical experiments on medium-large-scale problems from test collections such as CUTE show that the algorithm performs well. In addition, applying the method to image restoration and machine learning models further validate the effectiveness of the proposed method.

The deployment of medical AI agents represents a transformative paradigm in the healthcare sector. The medical AI agent ecosystem spans multiple disciplines, and its evaluation requires the collaboration and joint decision-making of experts from diverse fields. Due to limitations in domain-specific cognition and expertise, individual experts can typically provide only partial evaluations, making comprehensive assessment of all alternatives difficult. To address this, this study proposes a large-scale group decision-making model based on inclusion degree to overcome the dual challenges posed by numerous alternatives and large expert groups. First, the model classifies participants into parallel communities based on inclusion measure, thereby reducing group dimensionality and matching experts with appropriate alternatives. Next, a hierarchical consensus mechanism is constructed: type $\alpha$ consensus aimed at selecting the best alternative is used to manage conflicts within communities, and type $\gamma$ consensus aimed at ranking alternatives is used to reach a consensus across the entire group. Finally, the proposed model is validated through a study on the evaluation of medical AI agent deployment prioritization.

As competition among platform enterprises intensifies and development accelerates, introducing private labels and supporting small and medium-sized merchants have emerged as key strategic initiatives for platforms seeking new sources of profit growth. Accordingly, under a hybrid marketplace, this study employs the Spokes model to characterize an online sales system comprising a brand manufacturer, a small and medium-sized merchant, and an e-commerce platform. It further examines how factors such as product substitutability and perceived quality influence the platform's motivation to introduce private labels, as well as the subsequent effects on the decision-making and profitability of supply chain members. The results show that: (1) When product substitutability is high and consumers' perceived quality of the private label is low, introducing a private label increases both wholesale and retail prices of the brand manufacturer's and the small and medium-sized merchant's products. (2) Even when the seller's products are highly substitutable with the private label, introducing a lower-perceived-quality private label can still enhance sellers' profits. (3) When product substitutability is high and the private label's perceived quality is low, the platform still has strong incentives to introduce a private label, and doing so can achieve a win–win–win outcome for all three parties.

Aiming at the modeling problem of integer-valued time series data with limited value range, correlated counting events and hysteretic characteristics, this paper proposes a class of self-exciting hysteretic generalized binomial autoregressive models. Firstly, the strict stationarity and ergodicity of the model is proved, and the probabilistic and statistical properties of the model, such as conditional expectation, conditional variance and transition probability, are studied. Secondly, in both the cases where the threshold parameters and the initial values of the state indicators are known and unknown, the conditional least squares estimation and conditional maximum likelihood estimation methods of the model parameters are studied, and the asymptotic properties of the estimators are given. Among them, when the threshold parameters and the initial values of the state indicators are unknown, an algorithm capable of estimating the two threshold parameters is constructed. In addition, the testing method for the nonlinear structure of the model is studied. Subsequently, the effectiveness of the parameter estimation algorithms and the testing method for the nonlinear structure of the model are verified through simulation studies. Finally, the model is applied to the analysis of the number of precipitation days every two weeks in the Berleburg station of Germany and the Honolulu area of the United States. The results of the empirical analysis show that the model performs well in both fitting and prediction.

Joint mean and covariance modeling approaches with multivariate normal distributed errors are well-studied for longitudinal data. However, robust joint modeling approaches are less investigated when longitudinal data contains outliers or exhibits heavy-tails. In this paper, we establish the general form of joint mean and scale covariance models with multivariate Laplace distribution, sketch out the general maximum likelihood estimation methodology, and develop the general quasi-Newton algorithm. More specifically, we carry out three most popular decomposition methods as special cases, and establish respectively the corresponding joint model and estimation approach. Then strong consistency and asymptotic normality of the maximum likelihood estimator are proved. Moreover, a number of simulation studies are carried out to illustrate the modeling capability of these specific joint models, and real data is analyzed utilizing the specific decomposition method which performs best in simulation studies.

Air cargo is important in the global supply chain, and its demand is affected by many complex factors, which makes it difficult to forecast. In this study, an air cargo demand forecasting model is constructed based on TEI@I methodology, and the time series of cargo volume is decomposed using the complementary ensemble empirical mode decomposition method, the complexity and smoothness characteristics of each sub-series are analyzed. The appropriate seasonal autoregressive integrated moving average model or long- and short-term memory model is selected for forecasting. Besides, an expert system is introduced to deal with the irregular and important events, and finally the results of the various parts of the results are integrated to obtain the forecasting results. The results of the empirical study based on the data of Beijing Capital International Airport show that the forecasting effect of the model is better than that of other benchmark models, which provides a powerful support for the resource allocation and operation decision-making of air cargo enterprises.