In response to the rapid aging of society, implementing effective fertility support policies and delaying retirement policies has become a key focus of China’s future population policies. This paper, based on an overlapping generations general equilibrium model with endogenous fertility and endogenous retirement, assumes that the government allocates resources between fertility subsidies and pensions, and examines the impact of fertility subsidies on retirement decisions, pensions, and welfare in old age. The theoretical analysis shows that fertility subsidies not only increase the number of children born to residents but also encourage individuals to delay retirement. The impact of fertility subsidies on pensions follows an inverted U-shape, but they can increase old-age welfare, although this may reduce the welfare of retirees in the current period. Empirical results, based on OECD cross-national panel data, align with the theoretical findings. The results suggest that if the government faces resource constraints, shifting part of the resources from pensions to fertility support policies can both promote an increase in the birth rate and facilitate the implementation of delayed retirement policies. While this may affect the welfare of retirees in the current period, it will, in the long run, enhance the overall welfare level of future generations of elderly individuals. This study enriches the theoretical literature on fertility support and delayed retirement policies and provides valuable theoretical insights for the implementation of fertility support policies in China today.
Based on the data of China Family Panel Studies of 2016, 2018 and 2020, this paper validates the heterogeneous effects of air pollution on resident obesity and its impact mechanisms by using panel multivariate sorting models and ivoprobit models based on conditional mixed process. The results show that air pollution significantly promotes resident obesity; this conclusion remains valid after a series of robustness tests and mitigating the endogeneity problem. Specifically, air pollution promotes obesity more significantly for those in male group and outdoor working group, having a more evident impact on obesity among residents in rural and northern regions and the promotion effect is larger for residents of resource cities. Moreover, air pollution significantly contributes to resident obesity in mature and declining resource-based cities, but the promotion effect is not obvious in growing and regenerating ones. Further analysis reveals that residents’ age and emotion positively and negatively affect the relationship of air pollution on obesity, respectively; digital economy and green environment weakens the promotion effect. Further, we find that air pollution promotes resident obesity through health interference, activity inhibition, psychological pressure, as well as employment and income inhibition effect.
This paper combines the elastic network contraction method and the Diebold & Yilmaz method to measure the systemic network connectivity of China’s commodity futures market, which in turn reflects the systemic importance characteristics, and analyzes the degree of importance of each commodity futures variety in the system by combining the global network and different node perspectives. This paper finds that in static analysis, it is copper that lies at the dominant position of China’s commodity futures network, and in terms of major commodity categories, metal commodities play a higher degree of importance, followed by agricultural and chemical commodities, and in terms of different varieties, commodity futures tend to show bidirectional linkages between them of similar spillover strength, due to the association of industry chains and so on. In the dynamic analysis, the overall connectivity of the commodity futures system will relate to the point of major risk events related to the commodity market, which will further bring about changes in systemically important commodity varieties. In addition, factors such as changes in the supply and demand of the products themselves, industry chain linkages, and national industrial policies will also affect the turnover of systemically important varieties in the commodity futures system.
Under the strategy of main functional zones, development zones are important carriers for key development areas to play a role in economic growth. This paper focuses on distinguishing the levels and types of development zones, selects 893 districts and counties in key development areas in China from 2003 to 2020 as research samples, and uses the difference-in-difference method to re-evaluate the heterogeneous economic growth effects and driving mechanisms of various development zones at all levels. The results show that establishing national development zones and high-tech zones has a significant economic growth effect while establishing provincial development zones and economic development zones does not. This conclusion remained robust after the variable replacement, the sample change period, excluding policy interference and placebo testing. Heterogeneity analysis shows that the economic growth effect of key development areas varies according to the level of development zone, the type of development zone, the level of main functional zone, administrative divisions, geographical location, and development mode. Mechanism analysis finds that the differences in policy effect and agglomeration effect are the main channels leading to the heterogeneity of economic growth effect in key development areas. This paper further explains why establishing provincial-level development zones and economic development zones fails to promote regional economic growth from the perspective of public infrastructure. This study is of great significance for the in-depth implementation of the main functional zone system, giving full play to the role of key development areas as an important growth pole for economic growth, and optimizing the spatial development pattern of the territory.
In an economic environment of high uncertainty, reducing the risk of debt default and achieving a balance between “stable growth” and “risk prevention” are important safeguards to avoid systemic risks. This paper constructs a dynamic stochastic general equilibrium model with endogenous risk transfering of firms, and studies the increase of debt default risk caused by the information asymmetry between banks and firms. And we provide empirical evidence for the relevant results and theoretical mechanisms. Theoretical analysis reveals that aggressive investment decisions by small firms lead to higher debt default risk due to the presence of more severe information asymmetry. The risk transfer mechanism has a “risk accelerator” effect under uncertainty shocks, leading to asymmetric changes in debt default risk, which is amplified by higher risk assessments by banks. Under the risk-transfer mechanism, relevant policies should alleviate the information asymmetry problem from the perspective of credit supply. Empirical tests show that small firms have higher default risks when facing uncertainty shocks, and information asymmetry and risk-taking are important channels.
Discussions on the impact of IPOs on the ESG performance of companies in the same industry in the context of dual carbon goals and comprehensive registration system are helpful in understanding the relationship between capital market institutional development and the high-quality development of the real economy, with important practical significance. Empirical research based on Chinese stock market data from 2010 to 2022has found that IPOs tend to lower the ESG performance of companies in the same industry. The more intense the industry competition and the higher the similarity of products between companies in the same industry and newly listed companies, the greater the negative spillover effect. Additionally, the better the industry information environment, the weaker the negative spillover effect. These results indicate that the competitive pressure brought by newly listed companies will force companies to focus on their core business and weaken ESG investment. However, the incremental information provided during the IPO process can enhance the value of ESG investment and thereby promote corporate ESG investment. The weakening effect of competitive pressure outweighs the promoting effect of incremental information, leading IPOs to significantly lower the ESG performance of companies in the same industry. Further analysis shows that the negative spillover effect of IPOs on the ESG performance of companies is more prominent in non-state-owned enterprises, non-heavy polluting enterprises, companies with high management short-sightedness, and those with less financing constraints. While the comprehensive registration system enhances both competitive pressure and incremental information, the direct impact of competition compared to information, and the possibility of exchanges’ inquiries revealing negative information to exacerbate negative spillover, ultimately strengthen the negative spillover effects of the registration system. This study expands the research on the industry spillover effects of IPOs and deepens the understanding of corporate ESG investment decision-making from a capital market perspective.
Stock price information risk often refers to the asymmetric degree of important information affecting stock price among the informed and the uninformed traders. This paper constructs abnormal idiosyncratic volatility (AIV) based on the two most important information events of listed companies: earnings announcement and mergers and acquisitions (M&As), and takes it as a measure of information risk caused by informed trading and systematically discusses whether it is compensated for in expected stock returns. The empirical results show that: AIV is positively associated with the magnitude of informed return run-ups prior to earnings announcement and M&As, as well as short selling, insider trading and institutional trading during pre-earnings announcement and pre-M&As periods, which preliminarily provide evidence that AIV is related to information risk induced by informed traders; Using formal asset pricing tests, we find that stocks with higher AIV will have economically and statistically higher future return, that is, information risk captured by AIV is priced; AIV is a systematic risk factor that has a persistent impact on the stock price, which is distinct from post-earnings-announcement drift, post-M&As drift, idiosyncratic volatility anomaly, alternative quantity-based information risk measures, and mispricing measures.
Accurate prediction of stock volatility is crucial for investment and risk management, and it also forms the foundation of effective economic policy. This paper explores the effectiveness of interval data models in enhancing the accuracy of stock volatility predictions, covering indices from 19 developed and emerging markets. By comparing with GARCH, TGARCH, and the conditional autoregressive range (CARR) model proposed by Chou, the interval models demonstrated superior predictive performance. The study also employed data-driven methods to analyze the relationship between volatility and conditional ranges under non-normal distributions, conducting market analyses with data of various frequencies. Simultaneously, we explore the comparative performance of different models incorporating extreme value theory in tail risk prediction. Additionally, asset allocation research based on interval models found that their risk-adjusted returns significantly outperformed other models. These findings not only provide market participants with precise risk assessment tools but also offer robust methodological support for policymakers, aiding in the formulation of more effective policies.
Financial time series are characterized by nonlinearity and high noise. Many studies have developed predictive models using empirical mode decomposition and its derived algorithms, achieving relatively favorable prediction results. However, these studies often suffer from significant hindsight bias, leading to overly optimistic assessments of model performance. Although the stepwise decomposition method proposed later effectively avoids the problem of information leakage, the decomposed subsequences exhibit serious instability, resulting in poor prediction performance and even inferior to the undecomposed algorithm. This study conducted a thorough analysis of the shortcomings of existing methods, firstly discussing in detail the issue of prospective bias in the one-step decomposition prediction process, and verifying that the main reason for the destruction of effective information in the step-by-step decomposition prediction process is the influence of end effects; secondly, the research framework of the stepwise decomposition algorithm in practical applications has been improved, and a more optimal lagged empirical mode decomposition application method has been proposed, effectively alleviating the end effect problem and significantly improving the trend accuracy and stability of the prediction model; Finally, a trading system based on the LI-ELM-LOF machine learning hybrid model was proposed for trend prediction using extreme learning machine (ELM) under the lag decomposition algorithm (LI) and abnormal trend control using local outlier factor (LOF). Its effectiveness in multiple financial markets was verified through empirical research. In summary, the lagged decomposition system designed in this study achieves a balance between alleviating end effects and utilizing effective historical information. The trading results in eight futures market test sets demonstrate the potential of lagged decomposition algorithms in financial quantitative trading markets, providing a new approach for predicting non-stationary and nonlinear futures trends.
In the context of a global economic downturn in the post-pandemic era, the risk of automobile loan defaults in China has significantly increased, with default samples demonstrating an inherent imbalance. To enhance the sensitivity and interpretability of models for predicting default samples, this paper proposes an eXplainable heterogeneous weighted tree ensemble (XHWTE) model for automobile loan default risk prediction. The model incorporates a sample weighting mechanism tailored for imbalanced data and combines it with an F1-based threshold search algorithm, achieving post hoc interpretable analysis for heterogeneous weighted tree ensemble. Experiments conducted on domestic automobile loan default prediction datasets and the L&T automobile loan default dataset show that the XHWTE model improves F1 by 26.85% and 25.43%, respectively, compared to the random forest model. Furthermore, post hoc interpretative analysis at the variable level using weighted ensemble Shapley values identified key factors influencing automobile loan default risk predictions, such as loan-to-asset ratio and the loan staff. Comparative analyses, ablation studies, and statistical tests validate the significant advantages of the XHWTE model in terms of recall and F1. The proposed model not only significantly enhances predictive performance but also provides financial institutions with an actionable tool for effective risk management.
How to continuously and steadily monitor the development of objects in a complex information environment with a wide range of information sources and heterogeneous data is an important question in the field of evaluation and management. Based on the theory of generic comprehensive evaluation, this paper proposed a new class of comprehensive indices, namely multi-source random-aggregation-based indices, and discussed its influencing factors. In terms of index calculation, decision-maker preferences are incorporated into the calculation of benchmarks in the index by flexibly setting the distances between reference objects, which highlights the differences between benchmarks of multiple objects and responds to complex and diverse practice situations. Additionally, from the perspectives of horizontal comparison and vertical comparison, we presented two thoughts to calculate the multi-source random-aggregation-based index, and specifically gave three calculation rules. In the aspect of influencing factors analysis, through the large-scale simulations, it was found that setting the reference objects to 8-10 ensures the calculation accuracy and saves time cost. The benchmarks will decrease as the preference coefficient increases. The deviation calculation with horizontal thought can maximize the comparability among objects. Finally, through practical examples, it was shown that the multi-source random-aggregation-based index can continuously monitor the status of objects in increasingly complex evaluation environments, and at the same time has the characteristics of highlighting the differences between objects, which has both theoretical and practical value.
The government loan interest subsidies support enterprises in restoring production capacity. The government needs to consider the disaster situation of enterprises and determine an appropriate loan interest subsidy to enhance the efficiency of capital utilization and accelerate the speed of capacity recovery. This paper examines the design of loan interest subsidies by comparing subsidy rates and production capacity recovery efficiency across four types of contracts in an asymmetric information context: Single, Screening, Checking, and Penalty contracts. Furthermore, the paper analyzes the parameters that influence loan interest subsidy contracts. The results show that the government can adopt a combined contract with high (or low) loan interest subsidy and a strict (or relaxed) capacity recovery period. The disaster relief cost of the screening contract is consistently lower than that of the single contract. When the importance of emergency materials is low and there is a big gap in the quantity of materials, the checking contract for in-depth investigation of disaster situations in enterprises incurs the lowest cost. Under the checking contract, less affected enterprises might receive insufficient subsidies, hindering productivity recovery efficiency. The expanded analysis confirms the robustness of the main conclusions. This study offers decision support for the design of loan interest subsidy for government emergency relief loans.
Intelligent manufacturing can effectively optimize resource allocation, improve productivity and energy efficiency, and conserve energy. However, by altering corporate behavior and promoting scale expansion, it may also increase energy demand, thereby inducing an energy rebound effect. Based on the data of A-share listed enterprises in the manufacturing industry from 2010 to 2021, this paper investigates whether intelligent manufacturing induces the energy rebound effect by using the difference-in-difference-in-difference method and the energy rebound effect assessment model, considering the approval of intelligent manufacturing pilot demonstration projects as a quasi-natural experiment. The results of the study show that, first, intelligent manufacturing induces an energy rebound effect in the pilot enterprises, resulting in an increase of about 13.5 percent in their energy rebound effect. Second, intelligent manufacturing increases the demand for electricity through the substitution effect of electricity for other factors and the output effect of scale expansion, which in turn increases the energy rebound effect of the pilot enterprises. Third, intelligent manufacturing increases the energy rebound effect more significantly in energy-efficient enterprises, low energy-consuming enterprises, highly digitalised enterprises, state-owned enterprises, enterprises in the ten key industries of “Made in China 2025”, and enterprises in non-energy-consuming industries. This paper expands the mechanism and path of the energy rebound effect, enriches the assessment of the energy rebound effect and the research on energy saving and carbon reduction enabled by intelligent manufacturing, and provides theoretical basis and decision-making reference for the promotion of the synergistic development of digitalization and greening of the manufacturing industry.
Flight tests are essential to the R&D process for new aircraft. Numerous tests need to be assigned to appropriate flights to ensure the efficient and smooth execution. The current experience-based manual scheduling method is inefficient and cannot obtain high-quality schemes. To overcome these defects, this paper studies the flight test scheduling problem under complex constraints, establishes the mathematical model, and proves theoretical properties of the model. Meanwhile, this paper develops a large neighborhood search algorithm, and verifies the effectiveness of the model and the efficiency of the algorithm through comprehensive numerical experiments. Sensitivity analysis is conducted to compare the performance of the algorithm under different scenarios. The experimental results demonstrate that the developed algorithm can quickly obtain a high-quality flight test scheme, effectively improve scheduling efficiency, and provide a further scientific basis for the flight test scheduling.
Under the background that an escalating number of bond defaults in recent years, debt negotiation such as debt-equity swap to resolve strategic debt defaults by firms has emerged as a prevailing practice in the bond market. At the same time, the establishment of carbon market in China enables emission control firms to improve their expected cash flow by participating in carbon emission trading to alleviate their liquidity risks, and the uncertainty of carbon price fluctuations may increase the firm value. By incorporating carbon prices and carbon asset returns into a structural model of strategic debt servicing, the paper proposes a continuous-time theoretical model to study the impact mechanism of carbon emission reduction, carbon price and its volatility on the optimal timing of debt-equity swap and equity value of the firm quantitatively. The theoretical finding and the numerical analysis based on the data of emission control firms imply that the carbon price volatility has negative impacts on the optimal debt-equity swap trigger point and positive impacts on the equity value, respectively. The optimal trigger point would increase as both the bargaining power of shareholders and liquidation costs increase and the firm’s asset value volatility decreases, and each of the marginal impact of these three factors on the optimal trigger point increases. The numerical results further indicate that for a given carbon emission reduction, when it is positive, the carbon price is negatively correlated with the optimal debt-equity swap trigger point and positively correlated with the equity value. Meanwhile, the corresponding results are opposite when the given carbon emission reduction is negative. Moreover, carbon emission reductions have a positive correlation with the optimal trigger point and a positive correlation with equity value when carbon prices are given. The paper could deepen the understanding of the impacts of carbon prices on the policies of debt risk management for emission control firms, and provide significance suggestions and inspirations on how to improve the carbon market construction.
The intermittent consumption of renewable energy such as wind power has become one of the challenges in building a new type of power system. The integrated wind energy hydrogen storage and power generation system can effectively utilize the production and storage of hydrogen to increase the flexibility of wind power generation systems and improve the utilization efficiency of renewable energy. This article formulates a stochastic inventory model to address the decision-making problems related to production, storage, and energy conversion in the operational management of a wind-based integrated system of power generation and hydrogen storage. For this model, we show that its optimal cost function is “multi-modularity”, which is a structural property that is suitable for characterizing the energy efficiency of renewable energy generation systems coupled with storage. Based on complementary analysis, we characterize the optimal wind energy allocation strategy and optimal electric hydrogen energy complementary strategy for the integrated system, and analyze the sensitivity of these strategies relative to the system state parameters. The results indicate that, the conversion between electricity and hydrogen can greatly improve the energy utilization efficiency of the integrated system, and reduce its overall operating cost.
With the rapid development of the live-streaming sales business model and consumers’ preference for live-streaming shopping, it has become crucial for enterprises to plan their live-streaming channels. This paper considers a two-level supply chain consisting of a manufacturer and a retailer with an e-commerce channel and studies how the parties in the supply chain formulate strategies for introducing live-streaming channels, as well as the impact of introducing live-streaming channels on supply chain members and heterogeneous consumers. By establishing a game-theoretical model for four scenarios, including no live-streaming channels, only the retailer introducing live-streaming channels, only the manufacturer introducing live-streaming channels, and both the retailer and the manufacturer introducing live-streaming channels, this paper solves the optimal pricing and profit of both parties in the supply chain under each situation. Then, by comparing the profits under each scenario, the equilibrium live-streaming channel introduction decision of all parties in the supply chain is obtained. This study finds that: First, the retailer’s introduction of live-streaming channels alone does not necessarily increase her profit. Only when the consumers’ hassle cost of live-streaming shopping is small, the retailer will choose to introduce live-streaming channels. The introduction of live-streaming channels will increase both the manufacturer’s profit and consumer surplus. Second, the manufacturer’s introduction of live-streaming channels alone will increase his profit, and the introduction of live-streaming channels does not necessarily harm retailer profits. When the consumers’ hassle cost of live-streaming shopping is high, a win-win-win situation can be achieved for manufacturers, retailers, and consumers. Finally, when both retailers and manufacturers have the ability to operate live streaming channels, and the manufacturer has stronger live streaming capabilities, only one party will introduce live streaming channels in equilibrium: When the consumers’ hassle cost of live streaming shopping is very small or very large, only manufacturer opens live-streaming channel; when the cost is in the middle, only retailer opens live-streaming channel.
Rapid development of live streaming e-commerce has led us into an era where everything can be broadcasted. More and more products that rely on traditional offline channels are entering live streaming rooms, fully enjoying the demand dividend brought by massive traffic. This article focuses on the special scenario of jade live streaming and empirically studies why consumer purchase in live streaming e-commerce. Compared to other consumer goods, jade products are distinguished by high value, non-standardization, and information asymmetric between buyers and sellers. These characteristics make jade products seemingly not suitable for live streaming at all. However, in recent years, jade live streaming has gradually replaced offline channels and become the mainstream way of jade trading. Based on the information foraging theory and SOR (stimulus-organism-response) framework, this study investigates the impact of platform features (credibility, endorsement, and empowerment), anchor features (expertise, and influence), and live content features (social presence, vividness, and interactivity) on consumer purchase intention in jade live streaming. The potential impact mechanisms are examined from the perspectives of product discovery (perceived serendipity) and product evaluation (perceived diagnosticity and perceived symbolism). The research results indicate that platform endorsement is the most important factor affecting consumer jade purchases, with content vividness, content interactivity, social presence, and anchor expertise the next most important factors. Our findings not only enrich the literature on live streaming e-commerce, but also provide important insights for jade live streaming platforms, anchors, and merchants in planning live streaming activities and optimizing marketing strategies.
Aiming at the price competition and promotion cooperation between traditional e-commerce and content e-commerce, this paper establishes a noncooperative-cooperative biform game model to deeply explore how the two platforms jointly formulate operational strategies and distribute profits under the influence of consumers’ multi-attribute preference behavior. The research results show that: in the case of cooperative promotion between traditional e-commerce and content e-commerce, with the increase of promotion difficulty, the level of promotion efforts of content e-commerce, product pricing and profits of the two platforms have experienced a sharp decline in a relatively short range, and the commission paid by traditional e-commerce to content e-commerce first dropped abruptly and then slowly increased and gradually stabilized. Faced with the intensification of price competition, the two platforms choose to increase product prices and promotion investment can bring higher profits; With the increasing sensitivity of consumers to product promotion efforts, the commission paid by traditional e-commerce providers to content e-commerce first decreased slightly and then increased significantly, and the product prices, promotion efforts and profits of the two platforms have increased. The results of this paper can provide a theoretical basis for product pricing and operational decisions of e-commerce platforms enabled by content marketing.
The privacy security challenges faced by internet companies in the digital era are crucial and have become a focal point for government regulations. However, not all internet companies adhere strictly to security protocols. When subjected to external supervision, internet companies with different compliance incentives adopt varying strategies for privacy and data security disclosure, particularly in light of potential privacy breaches. This poses challenges for government regulation. To address this practical issue, this study constructs a signaling game model based on the privacy protection scenario of internet companies. It analyzes how the privacy security disclosure of internet companies influence the strategic choices of both players, considering policy support and punishment. Furthermore, the study demonstrates the rationality of the constructed equilibria using intuitive criteria and uncovers management insights behind their establishment. Additionally, the study conducts case analyses to examine the sensitivity of critical parameters and the development of equilibria. The findings indicate that government fines, while promoting the realization of separating equilibrium to some extent, can damage the government’s belief in maintaining privacy security if imposed excessively, thus hindering the policy goal of privacy protection. Moreover, when the privacy breach level is high, increased policy support from the government decreases its risk perception ability to some extent. Lastly, in cases where the privacy breach level is low, achieving secure data operations does not necessarily favor separating equilibrium over pooling equilibrium.
With increasing attention to corporate social responsibility (CSR), especially in the context of the rapidly developing platform economy, it is particularly important to study the recycling and sales mode choices of differentiated platforms, namely the primary platform (E1) and the secondhand recycling or sales platform (E2), from a CSR perspective focused on consumer surplus. Based on Stackelberg game theory, this paper constructs three decision models: a baseline model in which E1 does not participate in the secondhand market, and two models in which E1 is responsible for either recycling or selling secondhand products. This paper systematically analyzes how different resale strategies affect platform model selection, consumer surplus, environmental impact, and social welfare. The results show that when E1 fulfills CSR and sells secondhand products, consumer surplus, E1’s utility, and E2’s profit all increase compared to the case where E1 does not undertake CSR. Under both all- and partial-resale strategies, E1’s participation in recycling or sales improves its profitability but slightly reduces E2’s profit. Nevertheless, this scenario presents clear advantages in reducing environmental pollution and enhancing social welfare, while the baseline model is more favorable for improving consumer surplus. Moreover, with well-designed contractual mechanisms, effective coordination between platforms can be achieved. Further analysis reveals that fulfilling CSR in various forms consistently contributes to platform profit growth. When E2 assumes CSR, both platforms benefit. Under demand uncertainty, E1’s profit declines while its utility increases, and E2’s profit continues to rise.
Social network-data envelopment analysis method (SNA-DEA) is a data analysis approach that combines data envelopment analysis (DEA) with social network analysis (SNA). In SNA-DEA related studies, relational networks are commonly constructed based on reference relationships among decision-making units (DMUs) measured by DEA models in envelopment form. However, these studies ignore the fact that the reference relationships of inefficient DMUs may not uniquely determined, and constructing different relational networks and analyzing them based on different reference relationships can lead to inconsistent DMU ranking results. This inconsistency ultimately weakens the validity and reliability of the SNA-DEA method when applied in practice. To address this issue, this study proposes a method to construct an uncertain network by considering all possible reference relationships. The uncertain network-oriented HITS algorithm is then applied to analyze DMUs in the uncertain network, providing all possible ranking results and ranking probability for each DMU, yielding more reliable analytical outcomes. This approach holds significant theoretical and practical value for DMU evaluation, resource allocation optimization. Furthermore, positive and negative network construction methods are introduced, incorporating the maximization and minimization of the weight of reference relationships between DMUs. These methods allow for the generation of positive and negative rankings of the DMUs, respectively. These rankings offer decision-makers valuable insights for DMU ranking across different scenarios. Finally, the proposed approach is validated through the application of SNA-DEA method to examples.
Due to the severe resource conflict in multi-project systems, the classical single buffer monitoring model is not suitable for multi-project system management. In order to improve the efficiency of multi-project system management, this paper proposes a multi-project system collaborative management model for the project buffer and capacity constraint buffer based on sub-project criticality. Firstly, the importance of sub-project to system critical chain is analyzed, and the sub-project criticality coefficient is defined. Secondly, the capacity constraint buffer between sub-projects is determined based on sub-project criticality, then the capacity constraint buffer and the project buffer at the end of the sub-project are allocated and monitored collaboratively. Finally, the feasibility and optimization effect of the model is verified by system simulation examples. Simulation results illustrate that, the proposed collaborative buffer management approach can shorten the completion time and reduce the execution cost of the muti-project system compared with the single buffer management method.
Accompanied by the integration and optimization of factor endowments, the service-oriented manufacturing mode realizes the high degree of aggregation of manufacturing industry and service industry, thus generating the value capacity beyond the original industry, and empowering the economy and the society to realize the value-added. Firstly, based on the perspective of multi-dimensional value flow, this paper sorted out the internal driving force of the value-added of the service-oriented manufacturing model, and explained the external momentum of the value-added of the service-oriented manufacturing model with the help of the network embeddedness theory; secondly, based on the system dynamics and the causal mechanism, this paper constructed the network flow diagram of the service-oriented manufacturing at multi-levels, and formed the systematic expression of the value-added of the service-oriented manufacturing model; then, based on the simulation function calculation and model demonstration, system dynamics distribution and operation of service-oriented manufacturing are summarized; finally, based on the lean model idea and MDM matrix framework, this paper measured and captured the key drivers of value-added of service-oriented manufacturing model. We found that 1) the system of value-added service-based manufacturing model consists of service, integration and manufacturing level subsystems. The intertwining of different subsystems, levels and dimensions forms the micro, meso and macro expressions of service-based manufacturing value-added; 2) among the system dynamics, the structural embeddedness has a relatively balanced “bottoming mechanism”, the relational embeddedness has a fluctuating “linkage mechanism”, and the cognitive embeddedness has a stronger “amplification mechanism”, and the joint operation of multiple embedded activities has formed the extension of the value-added capacity of service-oriented manufacturing; 3) among the key drivers, the manufacturing layer contributes 28.6% of the value-added capacity, the service layer contributes 24.7%, and the fusion layer contributes 46.7%. Integration layer contributes 46.7% value-added, and there exists a value marginal effect logic of innovation > service > manufacturing in the service-oriented manufacturing model. The service-oriented manufacturing model under different frameworks and goal orientations has different key orientations.
In the case of violating the regularity assumption, the random utility model used in existing researches on unconstrained demand estimation of airline network cannot describe irrational customer choice behaviors. This research extends the traditional nonparametric discrete choice model to a partially-ranked preference lists based network model, and incorporates irrational index vector to develop a new framework that considers irrational customer strategic behavior in airline network. The EM algorithm is used for jointly estimating model parameters, and unconstrained demand calculation formulae considering both the spill and the recapture effects of historical irrational customer demand are proposed. Comparing with existing models, the network-based generalized stochastic preference model established in this paper performs better in the accuracy of demand estimation through numerical simulation, especially in the circumstances that irrational customer strategic behavior and network substitution effects are described.
