中国科学院数学与系统科学研究院期刊网
Home Browse Just accepted

Just accepted

Accepted, unedited articles published online and citable. The final edited and typeset version of record will appear in the future.
Please wait a minute...
  • Select all
    |
  • FANG Xin, ZHANG Chengyuan, CHAI Jian, WANG Shouyang
    Journal of Systems Science and Mathematical Sciences. https://doi.org/10.12341/jssms250013
    Accepted: 2025-04-03
    The volatility and nonlinear characteristics of time series have made modeling and prediction difficult and have attracted widespread attention from scholars. This study combines the decomposition and integration framework to achieve effective information extraction and modeling to improve prediction accuracy. Correspondingly, our proposed methodology involves four main steps: data decomposition via complete ensemble empirical mode decomposition with adaptive noise (CEEMDAN); component grouping via sample entropy (SE); prediction of the reorganized low-, mid-, and high-frequency component groups through persistence (PER), convolutional neural networks (CNN), gated recurrent unit (GRU), and ensemble prediction via weighted addition using ant lion optimization (ALO). Taking the hourly PM2.5 concentration of Xi’an as the sample, experimental results showed that our proposed hybrid decomposition-group-ensemble forecasting framework (i.e., ALO-CEEMDAN-SE-(PER-CNN-GRU)) significantly outperformed the benchmarks, and the final prediction error obtained the lowest value (2.53%). This validates the superiority of the decomposition integration framework with excellent neural network models for PM2.5 prediction.
  • ZHANG Zhixia, XU Ruliang, SHAO Bilin
    Journal of Systems Science and Mathematical Sciences. https://doi.org/10.12341/jssms240374
    Accepted: 2025-04-03
    Examining the dynamics of how group opinions evolve by considering the influence of the public environment and individual characteristics is crucial for accurately understanding the trends in public opinion propagation. It also offers theoretical guidance for managing the public opinion ecosystem. This paper analyzes the influence of group environment and individual characteristic factors in the formation of social individuals' opinions, and then constructs a heterogeneous opinion dynamic model, based on HK model, which account for environmental noise and individual opinion tolerance. Finally, it conducts with a simulation for analyzing the law of opinion evolution and the interpretability of the model to opinion evolution is verified by application in real-world network. The results demonstrate that individuals are prone to reach consensus with group views under the influence of high environmental noise, while low viewpoint tolerance thresholds make it more difficult to reach consensus and lead to fluctuating characteristics of opinion evolution. This research underscores the significant role of environmental factors and individual traits in opinion evolution, offering a fresh perspective for understanding and governing public opinion.
  • LI Minchan, LING Liwen, ZHANG Dabin
    Journal of Systems Science and Mathematical Sciences. https://doi.org/10.12341/jssms240567
    Accepted: 2025-04-03
    Financial market prices are affected by many factors and exhibit complex characteristics such as nonlinearity, non-stationarity, and high dimensions, making trend prediction very difficult. In order to further improve the accuracy of trend prediction, this paper proposes a model based on time series image coding and convolutional neural network optimized by genetic optimization,referred to as CMIE-GA-CNN (Combining Multi Image Encoding-Genetic Algorithm-CNN). First, from the perspective of feature engineering, aiming at the problem of difficult representation of high-dimensional characteristics of financial time series, the one-dimensional time series is encoded into two-dimensional images such as Gram graph, recurrence graph, Markov diagrams and line charts extract their high-dimensional invisible features. Secondly, the above images are combined to overcome the information loss problem of single-type image encoding. Finally, a genetic algorithm is introduced to optimize the parameters of the convolutional neural network to improve the performance of the prediction model. The combined graph is used as input to obtain the prediction results. In order to confirm the effectiveness of the proposed model, trend prediction and turning point prediction are performed on three groups of financial time series: the Dow Jones Industrial Average Index, the S&P 500 Index, and the U.S. Dollar Index, Compared with common prediction methods, single-class image encoding input and deep learning prediction methods without parameter optimization, the model in this paper improves the trend prediction accuracy and is more robust.
  • LI Xiaofei, LV Shuang
    Journal of Systems Science and Mathematical Sciences. https://doi.org/10.12341/jssms240582
    Accepted: 2025-04-03
    Against the backdrop of rapid development in e-commerce and retail, market competition continues to intensify. Accurate sales forecasting and dynamic pricing have become critical tools for retailers to enhance operational efficiency, optimize resource allocation, and strengthen competitive advantage within a complex market environment. Concurrently, the digital transformation and intelligent upgrading of the retail industry are driving companies from traditional experience-based management towards data-driven scientific decision-making. Particularly for perishable goods, swiftly responding to market changes and adjusting prices are crucial for inventory management, waste reduction, and improving sales efficiency. Based on this, this paper proposes a Retail Forecast Optimization and Commodity Classification Dynamic Pricing System, consisting of a high-precision forecasting subsystem and an intelligent pricing subsystem, aiming to achieve dual optimization of dynamic pricing and accurate sales forecasting. The high-precision forecasting subsystem constructs a hybrid model integrating Bidirectional Long Short-Term Memory (Bi-LSTM) and Bidirectional Gated Recurrent Unit (Bi-GRU), combined with Variational Mode Decomposition (VMD) technology, effectively capturing statistical patterns and trend fluctuations across different frequency sequences to realize high-accuracy multi-step forecasting of product sales and purchase prices; the intelligent pricing subsystem employs the IGWO-JAYA dynamic optimization algorithm to formulate dynamic pricing strategies for different categories, thereby maximizing revenue and achieving real-time market response. Results indicate that this system improves the prediction accuracy of product sales and purchase prices and can provide optimal pricing strategies for different categories of products, significantly enhancing retailers' adaptability to market changes and profit potential, supporting the intelligent development and efficiency improvement of the retail industry.
  • LI Qiang, SHI Huijun, HE Daojiang
    Journal of Systems Science and Mathematical Sciences. https://doi.org/10.12341/jssms240589
    Accepted: 2025-04-03
    The generalized log-Moyal (GLM) distribution, as a new class of heavy-tailed distribution, is closely related to the Moyal distribution, half-normal distribution, and $\chi^2$ distribution. In this paper, we adopt the objective Bayesian method to make statistical inferences for its parameters. Firstly, some noninformative priors, including the probability matching prior, Jeffreys prior, reference priors and maximum data information (MDI) prior, are derived, where it is shown that the reference prior is a probability matching one. Secondly, the posterior propriety based on the Jeffreys prior, reference prior and MDI prior is validated, where it is shown that the posterior distributions under the three priors are all proper. The simulation study demonstrates that the proposed Bayesian method offers several advantages over the maximum likelihood method. Finally, the proposed Bayesian method is applied to analyze a real data set.
  • HUANG Tingting, ZHANG Baoxue
    Journal of Systems Science and Mathematical Sciences. https://doi.org/10.12341/jssms240624
    Accepted: 2025-04-03
    Existing clustering methods for compositional data often fail to handle zero components. To address this limitation, we propose a clustering method specifically designed for compositional data with zero components. Considering the dependency of EM algorithm on the number of clusters and the initial values, a robust clustering algorithm with fixed initial values is developed by adding the penalty term of information entropy in the objective function. Numerical simulation experiments demonstrate that the algorithm can accurately and adaptively determine the number of clusters. Furthermore, compared to clustering methods based on α-transformation, the new method has better performance in recognizing the distribution patterns of compositional data. An analysis of household consumption structure survey data highlights the utility of the proposed method.
  • LEI Xiyang, MAO Chengyang, DAI Qianzhi, LIU Haoxiang
    Journal of Systems Science and Mathematical Sciences. https://doi.org/10.12341/jssms240674
    Accepted: 2025-04-03
    Improving energy and environmental efficiency is one of the effective means to achieve energy conservation and emission reduction in the transport system. When measuring the energy and environmental efficiency of China's transport system, it should be considered that China's transport system is a parallel structural system jointly composed of a freight subsystem and a passenger transport subsystem. At the same time, there is a game relationship among overall systems and it has not be considered in the majority of previous researches. Therefore, this paper combines the non-cooperative game with data envelopment analysis (DEA) to propose a non-cooperative game DEA method with considering the parallel structure feature of the transport system to measure the energy and environmental efficiency. Firstly, this paper designs an optimal algorithm for the efficiency evaluation on the overall system. The convergence of the algorithm and the uniqueness of the optimal solution have been proved. Then, on the basis of the leader-follower relationship among subsystems, a efficiency decomposition model is developed, and the existence on the solution of the efficiency decomposition model is further proved. Finally, the method of this paper is applied to measure the efficiency of transport systems in 30 provinces and municipalities in China in 2022. The results indicate that 1) The energy and environmental efficiency of China's transport system is generally low, and the optimal overall system efficiency of each region is not effective; 2) the energy and environmental efficiency of the passenger transport subsystem is better than that of the freight transport subsystem. 3) the overall system efficiency and passenger transport subsystem efficiency of China's transport systems in all regions show a pattern of "Central>Eastern>Western", and the freight transport subsystem efficiency shows a pattern of "Eastern>Western>Central". The results of the efficiency evaluation in this paper provide a direction and theoretical basis for the future improvement of the energy and environmental efficiency of China's transport system.
  • ZHAO Hui, ZHANG Yu, SHAO Mingyuan
    Journal of Systems Science and Mathematical Sciences. https://doi.org/10.12341/jssms240770
    Accepted: 2025-04-03
    Decentralized learning has gained increasing attention within the realm of big data distributed computing, owing to its merits in computational efficiency, data privacy protection and system stability. In the context of decentralized distributed learning, this paper proposes a sparse estimation method for expectile regression, leveraging the the asymmetric least squares loss and $\ell_1$ penalty. An ADMM-LAMM algorithm with a linear convergence rate is also outlined.Moreover, this paper establishes that the proposed estimator attains an approximately Oracle convergence rate and presents theoretical findings related to the recovery of the sparse support set. Lastly, numerical simulations and real data analysis are conducted to showcase the robustness and efficacy of the proposed methodology in handling heavy-tailed, heterogeneous high-dimensional data.
  • ZHANG Zhibin, CHIN Tachia, WANG Shouyang, CHIENG Weihua
    Journal of Systems Science and Mathematical Sciences. https://doi.org/10.12341/jssms240787
    Accepted: 2025-04-03
    The application of human-artificial intelligence (AI) integrated intelligent decision-making in emergency management has expanded significantly with advancements in AI, particularly in addressing complex scenarios such as flood disasters. However, subcultural cognitive differences critically influence the decision-making processes within such hybrid systems. Focusing on flood rescue operations, this study investigates the mechanisms through which subcultural disparities affect human-AI collaborative decision frameworks. While the integration of human intelligence (HI) and AI enhances system adaptability in uncertain environments, variations in decision priorities, risk assessments, and solution selection rooted in subcultural contexts may compromise the efficiency and effectiveness of cross-cultural emergency responses. By developing a simulation model, this research quantitatively analyzes the impact of subcultural differences on decision outcomes and proposes optimization strategies for human-AI systems. The findings contribute both theoretical frameworks and practical method to inform future cross-cultural emergency management research.
  • CHANG Ximing, KANG Zifan, FENG Ziyan, SUN Huijun
    Journal of Systems Science and Mathematical Sciences. https://doi.org/10.12341/jssms240817
    Accepted: 2025-04-03
    The rapid expansion of shared mobility services has introduced innovative solutions for urban transportation. Ride-hailing platforms, facilitated through user-friendly smartphone applications, seamlessly connect individual preferences with immediate vehicle availability. In carpooling services, passengers can share a ride in the same vehicle, setting their respective destinations as waypoints to increase vehicle utilization. This study proposes a carpooling and dispatching model for ride-hailing services based on a self-attention reinforcement learning network. Initially, a carpooling travel topology network is constructed, considering factors like passenger pick-up and drop-off times and locations. An on-demand algorithm is designed to identify ride-hailing orders suitable for carpooling. Subsequently, a self-attention reinforcement learning network is employed for order dispatching optimization. Through the implementation of policy gradient techniques for learning and training, the integration of masking methods ensures the efficacy of order dispatching. Leveraging the strengths of "offline training & online decision-making", the proposed strategy tackles the challenges of enhancing the real-time responsiveness of large-scale ride-hailing dispatching services. Finally, the real-world case study is conducted based on ride-hailing orders in Beijing, China. Results underscore the efficiency of the order dispatching algorithm in achieving near-optimal route selections while reaching a real-time demand response. Although carpooling slightly increases passenger waiting times, it significantly boosts ride-hailing operational efficiency, alleviates traffic congestion, and mitigates environmental pollutants.
  • ZHANG Faming, JIANG Yimeng, LUO Qian, LI Si
    Journal of Systems Science and Mathematical Sciences. https://doi.org/10.12341/jssms240904
    Accepted: 2025-04-03
    To improve the accuracy and reliability of multidimensional group evaluation information integration, this paper proposes a new method based on an improved power geometric Bonferroni mean operator. First, considering the differences in expert evaluations, the credibility of evaluation information is measured using cosine similarity and improved similarity at both the indicator and expert levels. Second, an HJ-CD model combining the advantages of Hamming and Jaccard similarities is constructed to quantify the proximity of expert evaluation information. Third, using the $P_2$GBM and GBM operators, the evaluation information and its credibility are integrated, and unreasonable information is corrected through forward correction, followed by layer-by-layer aggregation. Finally, the evaluation information is converted into real numbers using interval weighted midpoints to obtain the final evaluation result. The method is applied to the integrity evaluation of a coalbed methane compressor system, and comparative analysis demonstrates its feasibility and reasonableness.
  • LU Zhangxin, GUAN Yongqiang, XU Shiyang
    Journal of Systems Science and Mathematical Sciences. https://doi.org/10.12341/jssms240923
    Accepted: 2025-04-03
    Based on the leader-follower framework, this paper investigates the bipartite containment control problem of multi-agent systems with signed networks by using zero-sum game theory. Firstly, a two-person zero-sum game model is constructed, where two groups of leaders with conflicts of interest are regarded as two players in the game; the average distance between the final state of the follower agents and the leader agents are used as the incomes of the two players; each player chooses to connect with followers through $q$ edges at most in the game. Secondly, the properties of the game are analyzed from the perspective of interaction topology, and the results show that the number of connection edges between the two groups of leaders and followers in the optimal strategy of the game is always equal and the number is $q$. In addition, we analyze the case where each player selects only one follower agent to connect, it is concluded that when the connection edge signs of the two groups of leaders and followers are the same, the interaction topology corresponding to the circulant graph is an equilibrium topology. When the follower's interaction topology is structurally balanced graph, the payoffs of the two players are related to the number of agents in the two cells divided by the followers and the positive and negative signs of the connected edges. Finally, numerical simulation examples are given to illustrate the effectiveness of the theoretical results.
  • LI Yuanlu, LI Zhe, CHENG Libo, JIA Xiaoning
    Journal of Systems Science and Mathematical Sciences. https://doi.org/10.12341/jssms240947
    Accepted: 2025-04-03
    This paper considers the intrinsic directionality of rain streaks and the structural features of background images, and designs a single image rain removal model that incorporates multiple sparse priors in both vertical and horizontal directions. Initially, sparse priors are separately applied to the pixel domain and spatial domain of rain streaks. Subsequently, constraints are imposed using the transform domain and gradient domain of rain-free images to significantly preserve texture information of the image background while avoiding the loss and distortion of high-frequency image details. By integrating the aforementioned priors, a final convex minimization model is constructed and effectively solved using the Alternating Direction Method of Multipliers (ADMM). Moreover, rain streaks often have a certain angle with the vertical direction. This paper adopts a strategy of automatically recognizing the angle and rotating the image for processing. Experimental results demonstrate that the effectiveness of this algorithm in rain removal significantly surpasses that of other comparative algorithms, showing outstanding performance in both objective evaluation metrics and visual effects.
  • SHI Jilei, YIN Yilong, WEI Qi, SHAN Erfang
    Journal of Systems Science and Mathematical Sciences. https://doi.org/10.12341/jssms240960
    Accepted: 2025-04-03
    In cooperative games with graph structures, the player potential function method and the inclusion-exclusion decomposability property as two important methods can be used to study the Myerson value. The probabilistic Myerson value is a kind of generalization of Myerson value in cooperative games with generalized probabilistic graph structures, this paper shall extend the player potential function method and the inclusion-exclusion decomposability property to the cooperative games with generalized probabilistic graph structures, and define the probabilistic player potential function and probabilistic inclusion-exclusion decomposability property to explore their relationship with the probabilistic Myerson value and we give new characterizations and two new calculation methods of the probabilistic Myerson value. This study finds that the probabilistic player potential function method can give a more intuitive definition of the probability Myerson value from the perspective of marginal contribution, while the probabilistic inclusion-exclusion decomposability property can better reveal the internal relationship between the probabilistic Myerson value and probabilistic graph structure, so these conclusions has good theoretical value. Finally, an example of utility sharing of enterprise cooperative coalition is given to verify the rationality of the conclusion.
  • WEI Shuhao, YANG Yufei, YU Jiani, HUANG Hengzhen
    Journal of Systems Science and Mathematical Sciences. https://doi.org/10.12341/jssms241069
    Accepted: 2025-04-03
    An order-of-addition experiment is a scientific experiment where the experimental result is affected by changing the addition order of components. This paper considers the order-of-addition experiment with conditional effects, i.e., the situation where a specific component $k$ needs to be added before another specific component $l$. By using the general equivalence theorem, we first prove the optimality of the full design based on the order-of-addition model with conditional effects. It is found that the design structure of the order-of-addition with conditional effects under the pairwise-order model has a similar relationship with the structure of the order-of-addition orthogonal array of strength 3. Therefore, optimal fractional designs of the order-of-addition with conditional effects can be constructed based on order-of-addition orthogonal arrays of strength 3. The usefulness of the design and model is demonstrated using a single machine scheduling problem.
  • LU Jincheng, ZHUANG Hua, ZHOU Qin
    Journal of Systems Science and Mathematical Sciences. https://doi.org/10.12341/jssms241072
    Accepted: 2025-04-03
    The inverse data envelopment analysis (inverse DEA) method under the variable frontier assumption can effectively analyze the resource allocation and output optimization issues of decision-making units (DMUs) in future production periods, considering the impact of technological changes on the frontier. However, the existing methods only discuss the single case of frontier changes, ignoring the complexity of technological changes, which may lead to the diversity of frontier changes. To solve this problem, the article first reveals the reasons for the problems of the existing methods through in-depth analysis. Then, it proposes a variable frontier inverse DEA method under the assumption of technological diversity changes, and discusses the different scenarios of future production period frontier changes reflected by the proposed method and their related properties. Finally, through comparative analysis, the advantages of this method are demonstrated, and it is applied to the research on the planning of development goals for China's high-tech industries. The results show that the proposed method can effectively reflect the diversity of frontier changes and the potential changing trends of the frontier, and can provide reference for decision-makers to formulate relevant development plans.
  • ZHOU Yufeng, PENG Jing, BAI Yun
    Journal of Systems Science and Mathematical Sciences. https://doi.org/10.12341/jssms241019
    Accepted: 2025-03-17
    The study aims to develop a high-performance prediction model tailored to the blood collection and supply scenarios unique to China, considering its specific national conditions. It begins by analyzing seven factors:workdays, holidays, weekdays, months, seasons, winter and summer vacations, and the blood collection volume from the previous day. Statistical analyses confirm that all these factors significantly influence daily blood collection volumes. Subsequently, the study proposes a CNN-LSTM hybrid model that integrates Convolutional Neural Networks (CNN) and Long Short-Term Memory (LSTM) networks. The CNN component extracts periodic and local features from the data, while the LSTM component captures long-term temporal dependencies, enhancing feature representation capabilities. Experimental results demonstrate that the CNN-LSTM model outperforms other models, including CNN, LSTM, Generalized Regression Neural Network (GRNN), Back Propagation Neural Network (BPNN), Extreme Learning Machine (ELM), Seasonal Autoregressive Integrated Moving Average (SARIMA) and Linear Regression (LR). The CNN-LSTM model achieves the most comprehensive extraction of time series features across multiple factors and delivers the highest prediction accuracy. Specifically, its Normalized Mean Absolute Error (NMAE) and Normalized Root Mean Square Error (NRMSE) are reduced by up to 25.80% and 26.54%, respectively, while the coefficient of determination (R2) improves by up to 320.85%. The prediction results provide more precise decision-making references for blood collection and supply institutions, enabling better adjustment of collection plans and inventory management strategies.
  • CHEN Shengli, LI Xinru, LUO Menghua
    Journal of Systems Science and Mathematical Sciences. https://doi.org/10.12341/jssms240755
    Accepted: 2025-03-13
    As an important component of the modern economic system, digital finance has a crucial influence on the development of new quality productivity. Based on the panel data of 30 provinces in China from 2013 to 2022, this paper uses the entropy weight-TOPSIS method to measure the development level of new quality productivity at the provincial level, and analyzes the impact effect and mechanism of digital finance on new quality productivity through the two-way fixed effect model and the mediation effect model. The research finds that digital finance significantly promotes the development of new quality productivity, and this conclusion has passed the robustness test and endogeneity treatment. In the heterogeneity analysis, it is found that this promoting effect shows differences in different regions, different innovation capabilities and different degrees of enterprise agglomeration, presenting a pattern of "Central > Northeast > East > West", "High innovation capability > Low innovation capability", and "High degree of enterprise agglomeration > Low degree of enterprise agglomeration". The mechanism test finds that digital finance promotes new quality productivity through the positive effects of promoting the level of science and technology, improving the efficiency of resource allocation and optimizing the upgrading of industrial structure. The threshold effect analysis finds that when the level of innovation output crosses the threshold value in the process of digital finance influencing new quality productivity, the promoting effect of digital finance on new quality productivity weakens, and there is a marginal diminishing effect. Therefore, this paper discusses relevant policy suggestions, providing useful ideas for the formulation of policies on promoting the development of new quality productivity by digital finance.
  • LUO Suizhi, HU Sihuan, HE Xiaorong, CAI Mengsi
    Journal of Systems Science and Mathematical Sciences. https://doi.org/10.12341/jssms240901
    Accepted: 2025-03-13
    Online travel reviews encapsulate travelers' authentic experiences and serve as a crucial reference for potential tourists' decision-making. However, due to the vast number of reviews, it becomes challenging for potential tourists to browse each one and effectively extract valuable information. Moreover, existing research on travel itinerary recommendations often lacks sufficient sentiment analysis of online reviews, making it difficult to accurately reflect user needs. Considering that selecting a travel itinerary typically involves multiple constraints, this paper proposes a MACONT multi-attribute decision-making method based on picture fuzzy sets to recommend suitable travel itineraries for potential tourists. Firstly, we utilized the Octopus web scraping tool to collect travel itinerary review data from the Ctrip travel website and preprocessed the text using Jieba for word segmentation. Subsequently, the LDA topic model was employed to identify decision attributes and their weights for travel itineraries. Next, SnowNLP sentiment analysis was applied to extract the sentiment orientation of the reviews, quantifying them into picture fuzzy numbers. Then, integrating the MACONT multi-attribute decision-making method, a picture fuzzy MACONT decision model was constructed to achieve a comprehensive ranking of travel itineraries. Finally, a case study was conducted to validate the proposed method's rationality, and further sensitivity and comparative analyses were performed to demonstrate its effectiveness.
  • ZHAO Lili, LIU Zhenhao, YANG Xin
    Journal of Systems Science and Mathematical Sciences. https://doi.org/10.12341/jssms240940
    Accepted: 2025-03-13
    The issuance of green bonds is not only a key driver for enhancing enterprises' new quality productivity, but also an important means of deepening environmental responsibility practices. Based on the data from A-share listed enterprises from 2010 to 2022, this study employs a difference-in-differences model to analyze the impact of green bond issuance on enterprises' New Quality Productivity in China. The findings are as follows:(1) The issuance of green bonds significantly promotes the improvement of New Quality Productivity; (2) Mechanism analysis shows that the impact of green bond issuance on enterprises' New Quality Productivity lies in enhancing green innovation capability and reducing enterprise financing costs; (3) Heterogeneity analysis reveals that the promotion effect of green bond issuance on new quality productivity is more pronounced for non-state-owned enterprises, small and medium-sized enterprises, and high-pollution enterprises; (4) Further analysis indicates that public environmental attention and regional environmental regulations play a significant reverse moderating role in the process of improving enterprise new quality productivity. The results of this study provide empirical evidence for enhancing enterprises' New Quality Productivity, achieving green development transformation, and promoting high-quality development of enterprises.
  • ZHANG Peng, CHEN Wangxue
    Journal of Systems Science and Mathematical Sciences. https://doi.org/10.12341/jssms240884
    Accepted: 2025-03-10
    This paper explores the characteristics of optimal estimation for the population variance σ2 in a normal distribution N(μ2), where μ is known, utilizing balanced ranked set sampling(RSS). The theoretical findings indicate that the balanced RSS estimation of population variance exhibits greater efficiency compared to the estimation derived from simple random sampling (SRS). To enhance the efficiency of statistical inference, we propose a Fisher information maximization approach and a quasi-sufficient complete statistic framework for the RSS design. Furthermore, we investigate the optimal estimation of population variance and analyze its characteristics under these two methodologies. The numerical findings indicate that the RSS estimation of population variance utilizing Fisher information maximization, as well as the RSS estimation based on quasi-sufficient complete statistics, exhibit superior efficiency compared to the balanced RSS estimation. Furthermore, the quasi-sufficient complete statistic RSS estimation for population variance demonstrates greater efficiency than that derived from Fisher information maximization. Furthermore, this paper explores the characteristics of optimal estimation for the population variance σ2 in a normal distribution N(μ2), where μ is unknown, utilizing balanced RSS. The numerical findings indicate that the balanced RSS estimation of population variance exhibits greater efficiency compared to the estimation derived from SRS when μ is unknown. The real data analysis is provided to illustrate the numerical findings.
  • GUO Xiaole, RAN Bo
    Journal of Systems Science and Mathematical Sciences. https://doi.org/10.12341/jssms240855
    Accepted: 2025-03-07
    This paper deals with robust optimality condition and duality theory for a class of minimax fractional semi-infinite optimization problems with uncertain data. By virtue of robust optimization, Dinkelbach technique, and robust type constraint qualification conditions, we first establish robust optimality conditions for this uncertain optimization problem. Then, we introduce a Mixed type robust dual problem for this uncertain optimization problem, and explore robust duality properties between them. As a special case, we investigate robust optimality conditions and sum of squares relaxation properties for minimax fractional semi-infinite optimization problems with sum of squares convex polynomial structures.
  • QIAN Wuyong, GUO Kaiyi, WANG Xuan, XU Hanrong
    Journal of Systems Science and Mathematical Sciences. https://doi.org/10.12341/jssms240811
    Accepted: 2025-03-06
    Vehicle routing problem in takeout delivery is characterized by dynamic order arrivals and the need for continuous updates on rider status. To address this challenge, a multi-objective dynamic optimization model maximizes the interests of customers, platforms, and riders while considering rider physical condition and road familiarity. A dynamic weights multi-objective heuristic algorithm adaptively adjusts the weights of different objectives based on real-time data, optimizing delivery paths dynamically. Results demonstrate superior performance compared to the Gurobi solver in key metrics such as order fulfillment time, rider idle time, and platform profit. This highlights the effectiveness of the method in handling the complexities of real-world takeout delivery operations. Analysis of dispatch strategies for different types of riders provides valuable insights for operational decision-making. In summary, this research offers a practical solution to enhance delivery efficiency and customer satisfaction while ensuring fair treatment of riders, contributing to improved operational strategies for takeout platforms.
  • FANG Mengen, LI Lanqiang
    Journal of Systems Science and Mathematical Sciences. https://doi.org/10.12341/jssms240975
    Accepted: 2025-03-06
    Cyclic codes, an important subset of linear codes, are widely utilized in communication systems, consumer electronics, and data storage systems owing to their efficient encoding and decoding algorithms. This study aims to investigate construction of optimal ternary cyclic codes with parameter,[3m-1,3m-1-2m,4], By examining the existence of the solutions to specific equations over $\mathbb{F}_{3^m}$,we have obtained two distinct classes of optimal ternary cyclic codes. Furthermore, it is proved that such codes constructed in this paper are not equivalent to all known results, indicating that our results are new and have not been studied by other scholars.
  • YU Xiaohui, LIU Di, CUI Qingru
    Journal of Systems Science and Mathematical Sciences. https://doi.org/10.12341/jssms241036
    Accepted: 2025-03-06
    Under the "dual-carbon" goal, advanced manufacturing enterprises take green innovation as the core of development and actively improve their competitiveness. However, the green innovation of advanced manufacturing industry is a green innovation system composed of the government, the public, and advanced manufacturing enterprises (referred to as "enterprises"). The overall green innovation efficiency of the industry should be improved through the effective synergistic development within the system. Therefore, a three-party evolutionary game model consisting of enterprises, the government and the public is developed to analyze the impacts of the government's external incentives (including incubation platforms, R\&D subsidies and tax incentives) and the public's green preferences on enterprises' green innovation strategies. The study finds that:the effects of the three kinds of government external incentives to promote green innovation are not the same, among which R\&D subsidies are more effective in promoting green innovation at the early stage of green innovation. When the public's green preference is increased to a certain degree, the government can no longer give any external incentives to the enterprises, and then the enterprises can realize spontaneous green innovation. In the premise of no external incentives from the government, if we want to realize the spontaneous green innovation of the enterprise, then the enterprise's Green innovation is not the higher the better. In contrast, there exists an optimal degree of green innovation, when enterprises can realize spontaneous green innovation with the lowest public green preference requirement.
  • HAN Yongsheng, QI Zhiquan, TIAN Yingjie
    Journal of Systems Science and Mathematical Sciences. https://doi.org/10.12341/jssmsJSSC-2024-0088
    Accepted: 2025-03-05
    Learning from Label Proportions (LLP) is a weakly labeled learning problem, where the instance-level label information is abstracted in the form of bags, that is, only the label proportion information of each bag is available. Consequently, LLP can be grouped into learning with bags community, where bags consisted of instances are related. Similar to typical classification, our aim is not only to learn a classifier to greatly recover the instance-level labels in training data, but also to generalize this label prediction to unseen data. However, due to the ambiguous or approximate property in statistic estimation and the existence of label noises, a more realistic situation for this learning framework is prone to conceive an interval-type proportion information, instead of real-valued proportions in LLP. Thus, for these universal scenarios, the standard LLP methods are failed to offer a satisfied label predictor. In this paper, we propose a new learning framework called Bounded Label Proportions (BLP) to tackle this puzzled problem. In addition, we perform a robust algorithm for BLP based on Random Forest (RF):BLPForest, which is naturally able to deal with multi-class and high dimensional problems. For the purpose of comparison, we divided our experiments into two parts. In the first part, we degenerated BLPForest into standard LLP problem, in order to verify the evolution between these two similar learning problems. Consequently, the results demonstrated BLPForest with a natural advantage even in the case of real-valued proportion information equipped, which mainly benefited from the application of RF algorithm. For the second part, we chose large datasets with multi-class and much higher dimensions. In a meantime, appropriate noise for proportion information in each bag was deliberately added. All experiments showed that BLPForest can yield the best accuracies in the most cases. Finally, we offered the corresponding discussion and necessary analysis.
  • SHAO Zhen, ZHU Guowei, YANG Changhui, ZHAO Wei, LI Fei, LIU Chen
    Journal of Systems Science and Mathematical Sciences. https://doi.org/10.12341/jssms240779
    Accepted: 2025-03-04
    Accurately predicting and analyzing the complex trends of road transport carbon emissions among regions is crucial for setting and optimizing carbon emission reduction targets and promoting the synergy of pollution reduction and carbon reduction in the road transport industry. Aiming at the multiple time-varying characteristics of road traffic carbon emissions, such as time-varying peaks and seasonal changes, as well as the tightly coupled spatial characteristics brought about by the interconnections of different regional transportation, this paper takes into account the differences in climate regionalization and economic geographic distribution, and constructs a multi spatio-temporal fusion interregional road traffic carbon emission prediction model, MTGCN. First of all, a time-dependent relationship between short-term fluctuations and long-term trends of carbon emissions is captured by the temporal feature extractor. On this basis, the static and dynamic adaptive map structure information is integrated, and the spatial feature extractor is used to explore the potential spatial dependence of inter-regional road traffic carbon emissions. Finally, the validity of the proposed model is verified based on the daily carbon emission data of the Yangtze River Delta region.
  • QU Tianyao, LIAO Xiou, JU Xiaohang
    Journal of Systems Science and Mathematical Sciences. https://doi.org/10.12341/jssms240890
    Accepted: 2025-03-03
    Model averaging is a hot issue in the field of statistics and econometrics. Many model averaging methods have been proposed and the statistical properties of corresponding weights have been verified. Multivariate model is an important statistical model, which is widely used in various fields. In this paper, the convergence rate of weights in the model average estimation is obtained in the sense of MMMA and MJMA based on the linear statistical model with multiple dependent variables, and the verification results also cover the convergence rate of the model average estimated weights in the case of single dependent variables. In addition, we verify the convergence rate of weights by corresponding numerical simulation.
  • FENG Zhong-wei, REN Yu-hang, TAN Chun-qiao
    Journal of Systems Science and Mathematical Sciences. https://doi.org/10.12341/jssms241006
    Accepted: 2025-03-03
    This paper considers a two-period dynamic system with a proprietary brand manufacturer (PCM), an original equipment manufacturer (OEM) and strategic consumers, where PCM produces end products and proprietary components (PCs), decides whether and when to provide OEM with PCs, and determines when to enter the end market. The dynamic game models are constructed to explore the effects of product quality differentiation, purchasing behavior of strategic consumers, and bargaining power on the choice of PCM's coopetition strategies. The results show that:1) The strategy choice of PCM mainly depends on product quality differences when PCM has the independent pricing right for PCs. When OEM's product quality is low, PCM monopolizes the end market. When OEM produces high-quality products, PCM will provide OEM with PCs in the second period and enter the end market in both periods if the product quality differentiation is low, while PCM will provide OEM with PCs in the first period enter the end market in the second period if the product quality differentiation is high. 2) When PCM and OEM bargain over the wholesale price of PCs, providing components in the second period and entering the end market in both periods is PCM's inferior strategy. Whether to provide OEM with PCs in the first period and enter the end market in the second period depends on bargaining power, consumer patience, and product quality differentiation. 3) Compared to PCM's autonomous pricing situation, PCM is hurt by bargaining. However, if OEM is willing to redesign the profit-sharing mechanism, bargaining can achieve Pareto improvement in both parties' profits.
  • YANG Peng
    Journal of Systems Science and Mathematical Sciences. https://doi.org/10.12341/jssms240370
    Accepted: 2025-02-28
    This paper studies the reinsurance contract formulation problem under the game between two competitive reinsurers and one insurer. The insurer is engaged in two kinds of dependent insurance business, and the dependence is characterized by the number of claims and the amount of claims. The insurer signs reinsurance contracts with different reinsurers for the two kinds of insurance business. For these two kinds of insurance business, we don't restriction the reinsurance type, and the specific form of the reinsurance is determined by solving the stochastic optimization problem considered by the insurer and two reinsurers. Through relative performance, we quantify the competition between the two reinsurers, and establish a non-zero-sum stochastic differential game between them, and then establish a leader-follower stochastic differential game between the insurer and the two reinsurers. Under the mean-variance criterion, the explicit optimal reinsurance contract, i.e., the insurer's optimal claim risk sharing strategy and the two reinsurers' optimal reinsurance pricing strategy, is obtained by using stochastic analysis and stochastic control technology. Finally, the influence of model parameters on the optimal reinsurance contract is explored through numerical experiments.
  • YU Jinwei, ZHU Qi, MI Ruohan
    Journal of Systems Science and Mathematical Sciences. https://doi.org/10.12341/jssms240688
    Accepted: 2025-02-28
    By combining radial-based neural networks and event-triggered control strategies, the distributed formation control problem for nonlinear multi-robot systems with uncertain models is studied. And the conditions for triggering controller updates based on state information are provided. This control protocol can quickly achieve the formation tracking control objective by communicating with neighboring robots. The convergence of the system is independent of the initial states of the robots, effectively reducing the update frequency of the system's controller and the system's resource consumption. By using Lyapunov stability theory, it is proven that under the proposed protocol, the multi-robot system with uncertainties can achieve formation tracking without Zeno behavior. Finally, simulation examples are used to verify the feasibility of the theoretical results.
  • GUO Feng, HE Liang, SUN Xiangkai
    Journal of Systems Science and Mathematical Sciences. https://doi.org/10.12341/jssms240783
    Accepted: 2025-02-27
    This paper deals with a Tikhonov regularized primal-dual dynamical system featuring variable mass for solving convex optimization problems with linear equality constraints. Under suitable assumptions and through the use of energy functions, the convergence rates are first established for the primal-dual gap, the residual of the objective function, the feasibility measure, the velocity vector, and the gradient norm of the objective function along the trajectories. Then, the strong convergence of the primal trajectory of the dynamical system towards the minimal norm solution of the linear equality constrained convex optimization is demonstrated. Moreover, numerical experiments are conducted to illustrate the obtained results.
  • LAI Qinfei, WU Xianqing, WANG Zheyu, HE Xiongxiong
    Journal of Systems Science and Mathematical Sciences. https://doi.org/10.12341/jssms240524
    Accepted: 2025-02-24
    In practice, overhead cranes usually suffer from double pendulum effects which make the control of the crane systems more complex. However, for most existing traditional control methods, the people often not taken into account this double pendulum phenomenon. In addition, some trajectory planning methods have been proposed to improve work efficiency for crane systems, but the rope length in these methods is constant. To address these problems, this paper proposes a novel trajectory planning method with lifting/lowering operation for double-pendulum overhead crane. Specifically, to improve the efficiency and security of the transportation process, the trajectory is designed into three phases (acceleration, constant speed and deceleration). For each stage, the desired swing angle curve is directly constructed according to the requirements of the swing angle constraint and the zero residual swing angle, and the acceleration trajectory of the trolley is further obtained through the analysis of the dynamic equation of the double pendulum system. Then, the optimization mechanism is introduced, and the objective function about the transportation time and the maximum swing angle is constructed, and the trajectory planning problem is transformed into an optimization problem of the objective function. Finally, the simulation results are shown to verify the effectiveness of the proposed trajectory planning method.
  • WANG Chenbo, JI Zhijian
    Journal of Systems Science and Mathematical Sciences. https://doi.org/10.12341/jssms240557
    Accepted: 2025-02-24
    In this paper, the controllability of signed multi-agent networks based on the consensus protocol is studied from the perspective of topological structure. Firstly, based on the eigenvectors of Laplacian matrix and the leader-follower structure, the necessary and sufficient algebraic condition for the controllability of undirected signed topologies is obtained. According to the condition, for the composite topologies obtained by connecting two sub-topologies, two methods are proposed to construct controllable composite topologies by connecting the controllable sub-topologies. In addition, based on uncontrollable undirected signed topologies, the same sign double controllability destructive nodes (SSDCDN) and inverse sign double controllability destructive nodes (ISDCDN) are defined for the first time. By analyzing the characteristics of these nodes, the necessary and sufficient condition for the controllability on multi-leader undirected signed topological graphs is obtained. Finally, on the basis of the existing results, the design methods of two special types of uncontrollable signed sub-topologies connected with controllable signed sub-topologies to form controllable composite signed topologies are proposed.
  • QU Yunchao, CHANG Junbi, WU Jianjun, LEE Der-Horng
    Journal of Systems Science and Mathematical Sciences. https://doi.org/10.12341/jssms240668
    Accepted: 2025-02-24
    In recent years, the frequent occurrence of emergencies worldwide has not only severely impacted the socio-economy but also led to a large number of casualties. Emergency resource allocation is a key aspect of emergency management. When there is a shortage of supplies, transferring materials between disaster sites can improve the delivery efficiency and ensure the material needs of disaster victims. However, current research on dynamic emergency material allocation rarely considers the transfer of non-consumable materials between disaster sites, resulting in low allocation efficiency, insufficient flexibility, and lack of fairness. Therefore, this paper focuses on a scenario involving multiple supply points, multiple disaster sites, and various types of emergency materials. It establishes a dynamic allocation model that considers the transfer of materials between disaster sites, combining the allocation of materials from supply points to disaster sites and the transfer of materials between disaster sites. The model takes into account time-varying information such as supply and demand volumes, urgency of demand, transportation capacity, and road conditions, as well as the requirement that non-consumable materials must meet a certain service duration. It aims to formulate an emergency material allocation plan with efficiency and fairness as objectives, thereby improving allocation efficiency and optimizing the allocation of emergency resources. Through the solution and analysis of case scenarios, the model's effectiveness and the rationality of the allocation plan are verified in terms of emergency material scheduling efficiency, fairness, and the efficiency of considering material transfer between disaster sites.
  • SUN Chengyuan, WANG Xuesong, CHENG Yuhu
    Journal of Systems Science and Mathematical Sciences. https://doi.org/10.12341/jssms240662
    Accepted: 2025-02-21
    The existing quality-related fault diagnosis methods fail to reveal the intrinsic relationship between faults and quality due to the increasing complexity of industrial systems, and they also do not consider system dynamics thoroughly, leading to false alarms. False alarms lead to unnecessary maintenance and affect production efficiency, which will increase equipment costs and waste human resources. This paper proposes a quality-related interval fault diagnosis method based on multilevel decomposition to address this problem. Firstly, the method takes the nonlinear relationship between quality data and process data into full consideration and constructs the data model using multilevel decomposition strategy. Secondly, high-order discrete statistics are utilized to detect the system state, and a quality-related fault detection scheme is designed. Further, the interval dynamic fault diagnosis results are given by analyzing separation trajectories of fault samples and normal ones. Finally, the effectiveness of the method in this paper is verified based on the Tennessee Eastman platform and the wind turbine system.
  • DONG Minghua, CHU Chengpei, WANG Jianli
    Journal of Systems Science and Mathematical Sciences. https://doi.org/10.12341/jssms240686
    Accepted: 2025-02-11
    The increasing occurrence of extreme climate events poses a major threat to regional financial systems. Thus, exploring its impact mechanism is vital for preventing systemic financial risks. In this study, a comprehensive index analysis method and a DY variance decomposition matrix are used to calculate the regional financial risk volatility spillover values of 31 Chinese provinces and cities from 2014-2022. Three climate proxy indicators are constructed for a balanced panel regression to measure their impact on those spillover values. The findings show that extremely high temperatures have a significantly positive association with spillover, while extremely low temperatures and extreme precipitation have no significant effect. Analyzing regional heterogeneity reveals that regions like Inner Mongolia, Tibet, and Northwest China, with lower economic development levels and more homogeneous industrial structures, are more vulnerable to spillover from extreme climate shocks. Moreover, regional carbon emission intensity positively impacts the relationship between extreme high temperatures and spillover. Strengthening regional carbon reduction mechanisms is key to mitigating the effects of extreme high temperatures on spillover. In conclusion, enhancing regional governments' awareness to prevent extreme climate shocks and strengthening carbon reduction mechanisms are crucial for addressing risks from extreme climate events.
  • KANG Jijia, YANG Xiaoguang
    Journal of Systems Science and Mathematical Sciences. https://doi.org/10.12341/jssms241052
    Accepted: 2025-02-10
    Using ESG rating data of Sino-Securities Index Information Service from 2009 to 2020, this paper examines the impact of listed companies' ESG rating on the level of stock price, financial and operational risk in the next year. The study finds that better ESG rating has a significant inhibitory effect on all three risk levels of enterprises in the next year. Specifically, for the risk of stock price crash, ESG rating higher than the benchmark level, as a strong market signal, has a more significant reduction in the risk level of stock price crash. The trading volume of individual stocks, which reflects the attention of investors, has an intermediary effect on ESG to reduce the risk of enterprise stock price crash. ESG of large-scale enterprises that occupy an important position in the market and attract more attention from investors has a stronger inhibitory effect on the risk of stock price crash; In addition, the negative relationship between ESG and the risk of stock price crash is more significant after the implementation of the Environmental Protection Law. For financial risk, ESG has a marginal diminishing effect on reducing corporate financial risk, and the improvement of ESG rating from low to medium can improve the level of corporate financial risk. At the same time, enterprises' voluntary disclosure of non-financial information has a moderating effect on ESG rating to reduce corporate financial risks, and enterprises' voluntary disclosure strengthens the inhibitory effect of ESG on financial risks. For operational risk, ESG rating has a marginal diminishing effect on reducing operational risk; At the same time, the nature of equity has a moderating effect on the reduction of operating risks by ESG rating. Compared with private enterprises, ESG has a stronger inhibition effect on the operation risk of state-owned enterprises. Finally, the sub-sample heterogeneity test results based on the length of enterprise life in this paper show that the inhibitory effect of ESG rating on risk is stronger for enterprises with a long establishment age, but weaker for enterprises with a short establishment age.
  • PENG Yijian, TIAN Mengxin, JU Yuanyuan, WU Liucang
    Journal of Systems Science and Mathematical Sciences. https://doi.org/10.12341/jssms240359
    Accepted: 2025-01-20
    With the continuous development of Internet technology, data streams have been attracted wide attention. However, outlier may adversely affect the statistical inference of these data streams. So it is very important to study effective outlier detection methods. Because of the real-time characteristics, traditional outlier detection methods for data streams have been encountered challenges. To address this challenge, an online outlier detection method suitable for data streams are proposed in this paper. Firstly, the sample mean function of the data streams are updated online, and the principal component scores are updated to obtain the least trimmed scores set and evaluate the robustness of its mean estimator. Secondly, threshold rules are constructed based on the asymptotic distribution of distance to detect outlier, and one-step reweighting procedure is presented to control the false positive rate of outlier detection. Finally, the rationality and validity of the proposed method are verified by simulation and example analysis.
  • DENG Xin, YANG Yingxue, TANG Liping
    Journal of Systems Science and Mathematical Sciences. https://doi.org/10.12341/jssms240801
    Accepted: 2025-01-20
    In this paper, we introduce an approximate strong Karush-Kuhn-Tucker (ASKKT) conditions in response to smooth multiobjective fractional programming problems. In such problems, we obtained ASKKT-type necessary and sufficient optimality conditions at efficient solution. We also give the sufficient condition of ASKKT condition for Geoffrion properly efficient solution. Furthermore, we give the relation between ASKKT and classical SKKT conditions.