We conduct an empirical analysis of Shanghai-Hong Kong Stock Connect to reveal the dynamic impacts of stock connect trading activity on the stock pool's Amihud illiquidity proxy, index return, and CNY-HKD exchange rate. From pairwise conditional g causality analysis, we note a mutual significant causal connection between northbound net buying volume and Shanghai stock exchange return on all frequency levels. Meanwhile, we find a significant causal impact on the Shanghai portfolio's liquidity from northbound net buying volume. And there is a significant causal impact from the southbound net buying volume on Hang Seng Index return. Both are significant at the low-frequency level. In particular, northbound trading activity stimulates the Shanghai portfolio's liquidity in the low trading activity regime from the threshold VAR analysis. In robust analysis, we find similar significant dynamic causal connection and stimulation effects for the northbound trades when replacing Amihud illiquidity with the turnover rate. The result might relate to the investment behaviors looking for opportunity in the low trading activity regime. In contrast, the investors' beliefs may vary in the high trading activity regime, which weakens the connection between trading activities and other factors like liquidity.

Due to information asymmetry and strategic innovation, firms often encounter challenges related to insufficient driving forces and low-quality innovation outcomes. Analysts always act as information intermediaries who help foster the advancement of corporate innovation activities and the conversion of innovation output. This study examines the impact of analyst coverage and forecasting bias on corporate innovation, employing data from China A-shared listed firms spanning the period 2007 to 2019. We measure corporate innovation from two perspectives: Input and output. Specifically, we use the ratio of research and development (R&D) expenditure to sales as a proxy for the innovation input and the number of patent citations excluding self-citations to measure innovation output. We find that analyst coverage promotes corporate innovation, which is consistent with the "bright" side of analyst coverage. However, the positive effect of analyst coverage hinges on effectively transmitting and disclosing accurate information to investors in the capital market. Based on this, analysts' forecasting bias includes forecasting dispersion and optimism bias. We find evidence that an increase in analysts' forecast dispersion leads to a decrease in corporate innovation quality. Moreover, this paper presents a novel approach by employing the regression discontinuity method to examine the effect of analyst optimistic bias on firm innovation. The empirical findings reveal that overly optimistic forecasts by analysts exacerbate innovation quality. These analyses enrich the research on analyst coverage and corporate innovation, providing an empirical basis for improving the capital market with the help of analysts.

Based on the different premium volatility characteristics of various systematic factors in the A-share market, this paper constructs six representative high-frequency volatility prediction models that consider multiple complex risk structures. On this basis, a detailed comparative analysis of the differences in volatility characteristics among various factors is conducted, and the optimal prediction and early warning framework for the A-share market is proposed. Research shows that: 1) The volatility research results only for individual market indexes are not universally representative. 2) The fluctuation characteristics among different systematic factors and their respective optimal prediction model frameworks generally have significant differences, that is, there is no single fixed combination of model parameters. 3) Complex risk characteristics such as long memory, measurement errors, and high-frequency jump fluctuations obviously exist in the A-share market. The optimal forecast and early warning framework for the A-share market can be constructed by a combination of models that consider one or more of the above risk characteristics. The above conclusions have important practical reference value for the risk warning and prevention of the A-share market and the formulation of related policies.

To analyze the spatial influence mechanism of talent policy on population flow, this study compares the government work reports of 31 provinces between 2008 and 2020, and quantifies regional talent policies in nine aspects, including talent evaluation and incentives, utilizing a comprehensive, standardized, and continuous approach. Additionally, this paper develops a spatial econometric analysis model and expands on the conventional neighborhood, distance, and economic matrices by constructing a spatial weight matrix that reflects talent flow. The findings indicate that population movement exhibits spatial clustering patterns. The regional government's talent policy, primarily based on talent evaluation and incentives, positively influences population inflow. Moreover, during the implementation of talent policies, local governments demonstrate cooperative relationships. The inter-regional spillover effect between talent evaluation and talent incentives is significantly positive. In other words, a stronger local talent evaluation policy, along with robust talent incentives, encourages population inflow from neighboring provinces. However, this conclusion may vary in different regions and over time. Recently, the spatial spillover effect of population inflow and the impact of talent policies have not shown significant results. Additionally, the attractiveness of talent evaluation in the eastern region surpasses that of talent incentives, while the opposite holds true for the central and western regions. This study investigates the impact of local government talent policies on population inflow and its spatial spillover effect, offering theoretical support for intergovernmental cooperation.

This paper studies the regional differences, dynamic evolution and influencing factors of regional carbon emission intensity (CEI) in 262 cities and 5 regional urban agglomerations (UAs) in China. The Dagum Gini coefficient is used to analyze the intra-regional and inter-regional differences in carbon emissions, and the temporal evolution of the absolute differences of CEI among regions is analyzed by means of kernel density estimation (KDE). The paper provides an in-depth study on the spatial difference and temporal evolution of CEI in Chinese cities and major strategic regions. Through Moran index and LISA's test, the spatial correlation of carbon emission in prefecture-level cities is tested, and its spatial agglomeration characteristics are described. It is found that China's CEI is decreasing year by year, presenting a spatial pattern of plow in the south but high in the northq. Based on the calculation of carbon emission intensity at the urban level, this paper conducts LDMI factor decomposition research on carbon emission intensity at the national and key regions, and analyzes the impact of the impact factors on carbon emission intensity. The research results provide a path for China's green development at the city level and urban agglomeration level, and a theoretical support for different regions and cities to introduce emission and carbon reduction policies.

To analyze the influencing factors on the growth of economic and management talents in China's Sichuan Province, this paper constructs an analysis framework for the growth of economic and management talents, takes 340 economic management scholars in 53 higher education institutions in Sichuan Province as the research objects, uses CV analysis to organize their CVs and information, and constructs an evaluation index system combined with system science theory for the influencing factors from five dimensions: Educational experience, work experience, research experience, part-time experience, and award experience. Correlation analysis and structural equation model are used to systematically analyze the influencing factors on the growth of economic and management talents. The experimental results show that work experience, research experience, and award experience have a direct positive significant influence on talent growth; research experience and award experience play mediating roles in the influence of talent growth. This paper enriches the theoretical dimensions of this research field and explores the interactions among these factors. It also helps to improve the cultivation and development mode of economic and management talents in the western region. Furthermore, it provides guidance and reference for the role of talents in promoting economic growth, industrial upgrading, and sustainable development.

A three-dimensional boundary-spanning technology search model including search depth, scope and height is established, and a quantitative calculation method is proposed to dynamically describe an organisation's technology search behaviour and demand characteristics. Organisations are clustered by types as technical, comprehensive, or professional using k-means based on technology search behaviour. Recommendation strategies for various types of organisations are proposed based on this, and the search and supply libraries of each organisation are built by considering their type and search contents. The semantic similarity between patents in different libraries is calculated using a Word2Vec and TextRank model to achieve patent recommendations. An empirical study of the robotics field shows a recommendation accuracy of 0.751, and the accuracy of the technical, comprehensive, and professional types is 0.8282, 0.5389 and 0.7723, respectively. This study considers an organisation's dynamic search behaviour and makes class-based recommendations, with a low computational complexity and strong interpretability.

Motivated by the Bagging Partial Least Squares (Bagging PLS) and Principal Component Analysis (PCA) algorithms, a novel approach known as Principal Model Analysis (PMA) method is introduced in this paper. In the proposed PMA algorithm, the PCA and the Bagging PLS are combined. In this method, multiple PLS models are trained on sub-training sets, derived from the training set using the random sampling with replacement approach. The regression coefficients of all the sub-PLS models are fused in a joint regression coefficient matrix. The final projection direction is then estimated by performing the PCA on the joint regression coefficient matrix. Subsequently, the proposed PMA method is compared with other traditional dimension reduction methods, such as PLS, Bagging PLS, Linear discriminant analysis (LDA) and PLS-LDA. Experimental results on six public datasets demonstrate that our proposed method consistently outperforms other approaches in terms of classification performance and exhibits greater stability. Additionally, it is employed in the application of financial statement fraud identification. PMA and other five algorithms are utilized to financial statement fraud which concerned by the academic community, and the results indicate that the classification of PMA surpassed that of the other methods.

With the advancement of society and science and technology, the demand for detecting small objects in practical scenarios becomes stronger. Such objects are only represented by relatively small coverage of pixels, and the features are degraded severely after being extracted by a deep convolutional neural network, which is detrimental to the detection performance for small objects. Therefore, an intuitive solution is to increase the resolution of small objects by cropping the original image. In this paper, we propose a simple but effective object density map guided region localization module (DMGRL) to locate and crop the regions of interest where small objects may exist. Firstly, the density map of the objects is estimated by object density map estimation network, and then the coordinates of the small object regions are calculated; Secondly, the continuous differentiable affine transformation is utilized to crop these regions so that the detector with DMGRL can be trained end-to-end instead of two-stage training. Finally, the all prediction results of input image and cropped region images are merged together to output the final detection results by non maximum suppression (NMS). Extensive experiments demonstrate the superior performance of the detector incorporated DMGRL.

In the background of the green transformation of the economy and society, the ESG performance of enterprises has been paid more and more attention in the investment decision-making. However, previous studies have inadequately explored how the ESG performance affects corporate financing costs. Based on the information asymmetry theory, this paper analyzes the impact mechanism of ESG performance on corporate financing costs. Then, taking 1044 A-share listed companies in 2016–2020 as a sample, through the sorting and analysis of ESG report disclosure and rating data, the company's ESG performance indicators are obtained, and an empirical model is built to test the relationship between ESG performance and corporate financing costs. This paper constructs a panel regression model using ESG rating data and corporate financial data and finds that in the overall sample, the higher the ESG performance, the lower the equity financing cost; The higher the ESG performance, the lower the debt financing cost. In addition, it also discussed the moderating effect of enterprise scale and media attention on the impact of ESG performance on enterprise financing costs. The empirical results show that the influence of company size on ESG performance on financing costs has a moderating effect and a positive moderating effect.

During the COVID-19 pandemic, the international financial markets experienced severe turbulence. Under the background of "Made in China 2025", substantial entity enterprises have a large demand for non-ferrous metals. With the enhancement of financial attributes of non-ferrous metals, it is vital to prevent financial systemic risk contagion in the non-ferrous metal markets. In this article, the ensemble empirical mode decomposition method is used to decompose the prices of eight important non-ferrous metals futures, and then the dynamic DY risk spillover index model is established from the perspectives of long-term and short-term. The risk spillover between non-ferrous metals during the COVID-19 is quantitatively analyzed from different frequency domains. The study finds that in the long run, the risk spillover relationship between non-ferrous metals remained basically stable, and the change of it after the epidemic is slight. In the short run, the risk spillover relationship has different degrees of structural changes after the outbreak of the COVID-19 pandemic. The ensemble empirical mode decomposition method can distinguish the risk spillovers in different cycles, and help to formulate policies for preventing systemic risks in the non-ferrous metal markets according to the different length of terms.

This study investigates the impact of online-to-offline (O2O) platforms, such as Ele.me and Meituan, on offline sales in low-frequency-high-consumption industries, specifically a mid-to-high-end liquor distribution chain. Using data from 77 offline stores in Beijing collected during 2019–2022, the study employs a VAR model to analyze the relationship between offline sales and the use of O2O platforms. The results reveal a long-term equilibrium between the two, with most indicators showing a positive impact of O2O platforms on offline sales. The research provides valuable insights for low-frequency-high-consumption enterprises in making multi-channel decisions and quantifies the impact of O2O platforms on offline sales.

Cervical cancer is the fourth most common malignancy to strike a woman globally. If discovered early enough, it can be effectively treated. Although there is a chance of error owing to human error, the Pap smear is a good tool for first screening for cervical cancer. It also takes a lot of time and effort to complete. The aim of this study was to reduce the possibility of error by automating the process of classifying cervical cancer using Pap smear images. For the purpose of this study, dual convolution neural networks with LSTM were employed to classify images due to deep learning approaches inspire distinct features and powerful classifiers for many computer vision applications. The proposed deep learning model based on convolution neural networks (CNN) with the long short-term memory (LSTM) network is to learn features which give better recognition accuracy. The overall model is known as Smear-net. In which 'smear' indicates 'pap-smear cancer cells' and 'net' refers to neural network. The parameters such as, Accuracy, Precision, Recall, Accuracy, Sensitivity, and Specificity are used to validate the models. The proposed method provides the improved accuracy of 99.57 percentage for classification of the pap-smear cells. The proposed approaches demonstrate the effectiveness of our contributions by testing and comparing with the state-of-the-art techniques.

A method for internal participation in rescue decision-making of emergency volunteer teams considering psychological behavior is proposed to address the time sequence of rescue tasks. Firstly, the problem of multi-tasking and multi-operation within the emergency volunteer team is described. Secondly, considering that task leaders are influenced by behavioral and psychological factors in the evaluation, the required time for the job is used as a reference point, and the expected time that volunteers can complete the job is used as an attribute value. The task leader's prospect satisfaction value for each volunteer is calculated based on prospect theory, and the perceived utility values of disappointment theory and regret theory are calculated to measure the task leader's satisfaction with each volunteer. Furthermore, a multilayer coded genetic algorithm is used to construct an optimization model for emergency volunteer decision-making with the objective of maximizing the satisfaction value. Finally, the feasibility and effectiveness of this method are illustrated by an example analysis. The result shows that the efficiency of rescue tasks can be improved through decision optimization within the volunteer team.

This paper examines the data of A-share listed companies in China from 2002 to 2017, drawing on the theory of equal opportunity and market rules in M&A transactions. This paper investigates the correlation between changes in tender offer policy and M&A tendencies and performance. The findings suggest that following the policy shift and the adoption of market rules, companies that secure an exemption from the mandatory tender offer obligation not only exhibit stronger M&A tendencies but also improved long-term M&A performance. This indicates that market rules are more suitable for China and contribute to enhancing the efficiency of the M&A market. The paper also presents evidence of a moderating effect, demonstrating that exemptions from the mandatory tender offer obligation positively influence the relationship between policy change and M&A performance. Lastly, this paper finds that state-owned and large-scale firms tend to exhibit a higher degree of M&A tendencies.

This paper introduces the related concepts of the hybrid spherical-shaped dataset and proposes a new discriminant analysis method based on the spherical-shaped dataset (SDAM), then SDAM is further improved by the idea of the class cover and presents the nonlinear discriminant analysis method (NDAM). To demonstrate the effectiveness of these two methods, this work constructs seven hybrid spherical-shaped datasets and uses nine UCI datasets. Numerical experiments on these examples indicate that SDAM can preferably solve the discriminant problem for the hybrid sphericalshaped dataset, but this method does not always work well for real datasets; NDAM overcomes the drawbacks of SDAM and better solves the discriminative problem of real datasets. Besides, it has better stability.

This paper studies an online distributed optimization problem over multi-agent systems. In this problem, the goal of agents is to cooperatively minimize the sum of locally dynamic cost functions. Different from most existing works on distributed optimization, here we consider the case where the cost function is strongly pseudoconvex and real gradients of objective functions are not available. To handle this problem, an online zeroth-order stochastic optimization algorithm involving the single-point gradient estimator is proposed. Under the algorithm, each agent only has access to the information associated with its own cost function and the estimate of the gradient, and exchange local state information with its immediate neighbors via a time-varying digraph. The performance of the algorithm is measured by the expectation of dynamic regret. Under mild assumptions on graphs, we prove that if the cumulative deviation of minimizer sequence grows within a certain rate, then the expectation of dynamic regret grows sublinearly. Finally, a simulation example is given to illustrate the validity of our results.

To solve the problem that the traditional cloud model can't directly process the textual review information in the recommendation algorithm, this paper combines the merits of the cloud model in transforming qualitative and quantitative knowledge with the multi-granularity advantages of probabilistic linguistic term sets in representing uncertain information, and proposes a recommendation algorithm based on cloud model in probabilistic language environment. Initially, this paper quantifies the attributes in the review text based on the probabilistic linguistic term set. Subsequently, the maximum deviation method is used to determine the weight of each attribute in the evaluation information of the product to be recommended, and the comprehensive evaluation number and attribute weight are converted into the digital characteristic value of the cloud model by using the backward cloud generator. Finally, the products are recommended and sorted based on the digital characteristic value of the cloud model. The algorithm is applied to the recommendation of 10 hotels, and the results show that the method is effective and practical, enriching the application of cloud models in the recommendation field.

The land price in big cities draws much attention and discussion for its skyrocketing appreciation. Most researches are from the macro perspective due to data restriction. This paper aims to investigate the critical factors in the price formation process of a land auction, using the listing auction micro bidding-level data in Beijing from 2013 to 2018. We construct a model for the relationship between quitting price and land, bidder's characteristics, housing market conditions and competitive intensity (including private and public signals), then we use OLS for identification. We find that competitive intensity increases the quitting price by causing competition and interaction between bidders. More importantly, we find evidence of cheating behavior in the land market. Results show that bidders have higher quitting prices when they are in a joint venture, and when a central SOE developer or a top10 developer exist in the joint venture. We also find different behavior of developers in the short run and long run. Our research contributes to the literature of land auctions by analyzing the price formation process and developers' behavior. We also provide supporting evidence for the government to make adjustments of the auction system and identify the cheating developers.

Mountain cities are complex asymmetric dynamic network architectures, and the flight of UAVs in this environment is subject to various constraints, while efficiency is a crucial factor in the trajectory planning of police UAVs, which need to maintain high efficiency and safe flight paths between their starting and ending points, but the traditional trajectory planning method cannot meet the requirements of rapid maneuvering of police UAVs. To achieve this, a 3D terrain map is built, an objective function is established for the flight cost in the UAV trajectory planning process, and a planning algorithm called particle swarm optimization bat algorithm (PSOBA) is proposed. PSOBA combines the characteristics of the bat algorithm (BA) and the particle swarm optimization algorithm (PSO) to improve population diversity and resolve the delayed convergence issue in the last phases of BA. Simulation results show that PSOBA is more effective than BA, with a search time for the best solution that is approximately 20.43% shorter and a convergence value of the objective function that is approximately 38% smaller. PSOBA is also able to plan a quicker, shorter, and safer flight path compared to other trail planning algorithms that enhance the bat algorithm. These findings suggest that PSOBA is a powerful algorithm with potential application value in UAV trajectory planning control in the mobile intelligence era.Contribute to the service of public social security.

This study considers a nonlinear grey Bernoulli forecasting model with conformable fractional-order accumulation, abbreviated as CFNGBM$(1, 1, \lambda)$, to study the gross regional product in the Cheng-Yu area. The new model contains three nonlinear parameters, the power exponent $\gamma$, the conformable fractional-order $\alpha$ and the background value $\lambda$, which increase the adjustability and flexibility of the CFNGBM$(1, 1, \lambda)$ model. Nonlinear parameters are determined by the moth flame optimization algorithm, which minimizes the mean absolute prediction percentage error. The CFNGBM$(1, 1, \lambda)$ model is applied to the gross regional product of 16 cities in the Cheng-Yu area, which are Chongqing, Chengdu, Mianyang, Leshan, Zigong, Deyang, Meishan, Luzhou, Suining, Neijiang, Nanchong, Guang'an, Yibin, Ya'an, Dazhou and Ziyang. With data from 2013 to 2021, several grey models are established and results show that the new model has higher accuracy in most cases.

This paper considers sales mode selection issues for a two-echelon supply chain of perishable products involving a supplier and a retailer. Consumers' purchase desire is assumed to be positive correlation with the freshness of the product. First, the traditional sales mode is examined, in which the supplier and the retailer play a Stackelberg game. We propose a three-layer decision model for this situation, and obtain dynamic pricing strategies and selling cycle length. It is shown that the retailer has little motivation to order many perishable products so as to avoid a long selling cycle length. Second, the commission-charge mode is analyzed, in which the retailer declares its decision first. In this mode, we demonstrate that the perishable product will be on sale during the whole shelf life under a certain condition. The correlation between the sales price of each stage and the remaining shelf-life length is analyzed. Third, the superiority analysis for the two sales modes is conducted. We show the relation between the selling cycle lengths of the two modes. By our analysis, it is shown that both the supplier and the retailer gain more profits when the commission-charge mode is adopted and the commission rate locates in a certain open interval. Finally, a numerical illustration is presented to visualize the discussed models, and some supplements are made for the acquired conclusions by the illustration.

The probabilistic hesitant fuzzy multi-attribute group decision-making method introduces probability and hesitation into decision-making problems at the same time, which can improve the reliability and accuracy of decision-making results, and has become a research hotspots in recent years. However, there are still many problems, such as overly complex calculations and difficulty in obtaining probability data. Based on these, the paper proposes a multi-attribute group decision-making model based on probability hesitant fuzzy soft sets. Firstly, the definition of probabilistic hesitant fuzzy soft set is given. Then, based on soft set theory and probabilistic hesitant fuzzy set, the similarity measure of probabilistic hesitant fuzzy soft set is proposed, and the two measures are further combined. Finally, it is applied to the construction of multi-attribute group decision-making model, and the effectiveness and rationality of the model are verified by an example. The example shows that the new similarity calculation formula and algorithm model in this paper have higher accuracy, and the calculation process is more simple, it provides a feasible method for multi-attribute group decision making problems.