Search results

As intelligent technology advances, practical applications often involve data with multiple labels. Therefore, multi-label feature selection methods have attracted much attention to extract valuable information. However, current methods tend to lack interpretability when evaluating the relationship between different types of variables without considering the potential causal relationship.

To address the above problems, we propose an ensemble causal feature selection method based on mutual information and group fusion strategy (CMIFS) for multi-label data. First, the causal relationship between labels and features is analyzed by local causal structure learning, respectively, to obtain a causal feature set. Second, we eliminate false positive features from the obtained feature set using mutual information to improve the feature subset reliability. Eventually, we employ a group fusion strategy to fuse the obtained feature subsets from multiple data sub-space to enhance the stability of the results.

Experimental comparisons are performed on six datasets to validate that our proposal can enhance the interpretation and robustness of the model compared with other methods in different metrics. Furthermore, the statistical analyses further validate the effectiveness of our approach.

The present study makes a noteworthy contribution to proposing a causal feature selection approach based on mutual information to obtain an approximate optimal feature subset for multi-label data. Additionally, our proposal adopts the group fusion strategy to guarantee the robustness of the obtained feature subset.

This paper aims to propose an incremental graph indexing method based on probabilistic inferences in Bayesian network (BN) for approximate nearest neighbor search (ANNS) that adds unindexed queries into the graph index incrementally.

This paper first uses the attention mechanism based graph convolutional network to embed a social network into the low-dimensional vector space, which could improve the efficiency of graph index construction. To add the unindexed queries into the graph index incrementally, this study proposes to learn the rule-based BN from social interactions. Thus, the dependency relations of unindexed queries and their neighbors are represented, and the probabilistic inferences in BN are then performed.

Experimental results demonstrate that the proposed method improves the search precision by at least 5% and search efficiency by 10% compared to the state-of-the-art methods.

This paper proposes a novel method to construct the incremental graph index based on probabilistic inferences in BN, such that both indexed and unindexed queries in ANNS could be addressed efficiently.

Mega construction projects (MCPs), which play an important role in the economy, society and environment of a country, have developed rapidly in recent years. However, due to frequent social conflicts caused by the negative social impact of MCPs, social risk control has become a major challenge. Exploring the relationship between social risk factors and social risk from the perspective of risk evolution and identifying key factors contribute to social risk control; but few studies have paid enough attention to this. Therefore, this study aims to systematically analyze the impact of social risk factors on social risk based on a social risk evolution path.

This study proposed a social risk evolution path for MCPs explaining how social risk occurs and develops with the impact of social risk factors. To further analyze the impact quantitatively, a social risk analysis model combining structural equation model (SEM) with Bayesian network (BN) was developed. SEM was used to verify the relationship in the social risk evolution path. BN was applied to identify key social risk factors and predict the probabilities of social risk, quantitatively. The feasibility of the proposed model was verified by the case of water conservancy projects.

The results show that negative impact on residents’ living standards, public opinion advantage and emergency management ability were key social risk factors through sensitivity analysis. Then, scenario analysis simulated the risk probability results with the impact of different states of these key factors to obtain management strategies.

This study creatively proposes a social risk evolution path describing the dynamic interaction of the social risk and first applies the hybrid SEM–BN method in the social risk analysis for MCPs to explore effective risk control strategies. This study can facilitate the understanding of social risk from the perspective of risk evolution and provide decision-making support for the government coping with social risk in the implementation of MCPs.

The purpose of this paper is to explore a new grey relational analysis model to measure the coupling relationship between the indicators for the water environment status assessment. Meanwhile, the model deals with the problem that the changing of indicator order may result in the changing of the degree of grey relation.

The binary index submatrix of the sample matrix is given first. Then the product of the matrix and its own transpose is used to measure the characteristics of the index and the coupling relationship between the indicators. Thirdly, the grey relational coefficient is defined based on the matrix norm, and a grey coupling relational analysis model is proposed.

The paper provides a novel grey relational analysis model based on the norm of matrix. The properties, normalization, symmetry, relational order invariance to the multiplicative, are studied. The paper also shows that the model performs very well on the water environment status assessment in the eight cities along the Yangtze River.

The model in this paper has supplemented and improved the grey relational analysis theory for panel data.

Product returns are a major challenge for e-businesses as they involve huge logistical and operational costs. Therefore, it becomes crucial to predict returns in advance. This study aims to evaluate different genres of classifiers for product return chance prediction, and further optimizes the best performing model.

An e-commerce data set having categorical type attributes has been used for this study. Feature selection based on chi-square provides a selective features-set which is used as inputs for model building. Predictive models are attempted using individual classifiers, ensemble models and deep neural networks. For performance evaluation, 75:25 train/test split and 10-fold cross-validation strategies are used. To improve the predictability of the best performing classifier, hyperparameter tuning is performed using different optimization methods such as, random search, grid search, Bayesian approach and evolutionary models (genetic algorithm, differential evolution and particle swarm optimization).

A comparison of F1-scores revealed that the Bayesian approach outperformed all other optimization approaches in terms of accuracy. The predictability of the Bayesian-optimized model is further compared with that of other classifiers using experimental analysis. The Bayesian-optimized XGBoost model possessed superior performance, with accuracies of 77.80% and 70.35% for holdout and 10-fold cross-validation methods, respectively.

Given the anonymized data, the effects of individual attributes on outcomes could not be investigated in detail. The Bayesian-optimized predictive model may be used in decision support systems, enabling real-time prediction of returns and the implementation of preventive measures.

There are very few reported studies on predicting the chance of order return in e-businesses. To the best of the authors’ knowledge, this study is the first to compare different optimization methods and classifiers, demonstrating the superiority of the Bayesian-optimized XGBoost classification model for returns prediction.

The purpose of this study is to investigate how the implementation of digital banking services (mobile applications) by globally systemically important banks (G-SIBs) affects banks’ performance in the USA and Europe from 2005 to 2022.

The study employs advanced econometric methods to analyze the link between deposits and banking performance, utilizing linear regressions and multivariate Bayesian regressions.

Our results indicate that customer deposits positively impact a bank’s performance after the introduction of the mobile application feature of check deposits, whereas social risk negatively impacts banking financial performance. These findings support the hypothesis that technology implementation improves the profitability and growth of traditional banks.

While findings are robust econometrically in linear and Bayesian regressions, variables reflecting the digitalization of banks remain limited. For instance, the number of mobile users or the volume of digital transactions per bank since the implementation of the mobile app is not available.

In a rapidly growing technology and constantly changing customers behaviors, this research has practical implications from bankers’ perspective to continue the technological innovation efforts and from regulators’ perspective to strengthen requirements for the digital banking services.

We provide empirical evidence that including a banking app for smartphones’ users for remote banking services benefit the financial performance of banks. However, the social risk remains significant for banks in terms of customers' satisfaction, data privacy and cybersecurity.

This paper employs an innovative approach to create a mobile app “discriminatory” factor and examine the relationship between deposits and banks’ performance before and after the introduction of a mobile app for too-big-to-fail banks in Europe and the USA. Additionally, we consider the social risk component of the ESG score, as a bank’s decision to implement mobile applications and technology for its customers potentially affects social risks associated with customer satisfaction and technology usability.

Experiential learning is crucial in education, as it offers hands-on, practical experiences that enable individuals to develop their skills and knowledge more engagingly and interactively. In recent years, experiential learning has become a significant aspect of education. To provide academic scholars with a thorough roadmap for further investigation, this study aims to provide useful insights into the bibliometric and content analysis of experiential learning, including keywords, well-known authors, publications, nations and topics.

This research does a rigorous bibliometric analysis to give a thorough and visually instructional assessment of the evolution and advancement of the literature on experiential learning. Its fast development between 1976 and 2022 is meticulously tracked in the research. By using VOSviewer and Biblioshiny tools, the present study presents a concise overview of 507 records retrieved from the Scopus database using the keyword “Experiential Learning”, following the Preferred Reporting Items for Systematic Reviews and Meta-Analysis protocol. Deeper text mining was done using Python libraries “Pandas” and “Natural Language Toolkit” and regular expressions.

The findings reveal a surge in the number of publications on experiential learning and provide insights, particularly using the theory, context, characteristics, methodology analysis, supporting researchers and practitioners to understand learning better and provide perspectives for future research. Descriptive bibliometric analysis showed that most contributions are from the USA, the UK and Canada. In-depth content analysis revealed five clusters: developments in learning, management education, engineering curricula, organisational learning and knowledge management and entrepreneurship education. The keyword co-occurrence analysis enabled linkages between relevant fields of study and significant research domains. The most commonly used theories were: experiential learning theory, social learning theory, relational coordination theory, empowerment theory, feedback learning theory, effectuation theory and human capital theory.

This study uses information from the Scopus database to do a bibliometric analysis of experiential learning from 1976 to 2022. This study serves as a valuable resource for researchers in the field, helping them to position their work more explicitly within the existing literature and highlighting potential areas for future research. It does this by thoroughly analysing the literature on experiential learning using bibliometrics.

This research proposes a novel risk assessment model to elucidate the risk propagation process of industrial safety accidents triggered by natural disasters (Natech), identifies key factors influencing urban carrying capacity and mitigates uncertainties and subjectivity due to data scarcity in Natech risk assessment.

Utilizing disaster chain theory and Bayesian network (BN), we describe the cascading effects of Natechs, identifying critical nodes of urban system failure. Then we propose an urban carrying capacity assessment method using the coefficient of variation and cloud BN, constructing an indicator system for infrastructure, population and environmental carrying capacity. The model determines interval values of assessment indicators and weights missing data nodes using the coefficient of variation and the cloud model. A case study using data from the Pearl River Delta region validates the model.

(1) Urban development in the Pearl River Delta relies heavily on population carrying capacity. (2) The region’s social development model struggles to cope with rapid industrial growth. (3) There is a significant disparity in carrying capacity among cities, with some trends contrary to urban development. (4) The Cloud BN outperforms the classical Takagi-Sugeno (T-S) gate fuzzy method in describing real-world fuzzy and random situations.

The present research proposes a novel framework for evaluating the urban carrying capacity of industrial areas in the face of Natechs. By developing a BN risk assessment model that integrates cloud models, the research addresses the issue of scarce objective data and reduces the subjectivity inherent in previous studies that heavily relied on expert opinions. The results demonstrate that the proposed method outperforms the classical fuzzy BNs.

By introducing Stacking algorithm to solve the underfitting problem caused by insufficient data in traditional machine learning, this paper provides a new solution to the cold start problem of entrepreneurial borrowing risk control.

The authors introduce semi-supervised learning and integrated learning into the field of migration learning, and innovatively propose the Stacking model migration learning, which can independently train models on entrepreneurial borrowing credit data, and then use the migration strategy itself as the learning object, and use the Stacking algorithm to combine the prediction results of the source domain model and the target domain model.

The effectiveness of the two migration learning models is evaluated with real data from an entrepreneurial borrowing. The algorithmic performance of the Stacking-based model migration learning is further improved compared to the benchmark model without migration learning techniques, with the model area under curve value rising to 0.8. Comparing the two migration learning models reveals that the model-based migration learning approach performs better. The reason for this is that the sample-based migration learning approach only eliminates the noisy samples that are relatively less similar to the entrepreneurial borrowing data. However, the calculation of similarity and the weighing of similarity are subjective, and there is no unified judgment standard and operation method, so there is no guarantee that the retained traditional credit samples have the same sample distribution and feature structure as the entrepreneurial borrowing data.

From a practical standpoint, on the one hand, it provides a new solution to the cold start problem of entrepreneurial borrowing risk control. The small number of labeled high-quality samples cannot support the learning and deployment of big data risk control models, which is the cold start problem of the entrepreneurial borrowing risk control system. By extending the training sample set with auxiliary domain data through suitable migration learning methods, the prediction performance of the model can be improved to a certain extent and more generalized laws can be learned.

This paper introduces the thought method of migration learning to the entrepreneurial borrowing scenario, provides a new solution to the cold start problem of the entrepreneurial borrowing risk control system and verifies the feasibility and effectiveness of the migration learning method applied in the risk control field through empirical data.

The study examines the impact of behavioral biases, such as herd behavior, overconfidence and reactions to ESG News, on Socially Responsible Investing (SRI) decisions in the Indian context. Additionally, it explores gender differences in SRI decisions, thereby deepening the understanding of the factors shaping SRI choices and their implications for sustainable finance and gender-inclusive investment strategies.

The study employs Bayesian linear regression to analyze the impact of behavioral biases on SRI decisions among Indian investors since it accommodates uncertainties and integrates prior knowledge into the analysis. Posterior distributions are determined using the Markov chain Monte Carlo technique, ensuring robust and reliable results.

The presence of behavioral biases presents challenges and opportunities in the financial sector, hindering investors’ SRI engagement but offering valuable opportunities for targeted interventions. Peer advice and hot stocks strongly predict SRI engagement, indicating external influences. Investors reacting to extreme ESG events increasingly integrate sustainability into investment decisions. Gender differences reveal a greater inclination of women towards SRI in India.

The sample size was relatively small and restricted to a specific geographic region, which may limit the generalizability of the findings to other areas. While efforts were made to select a diverse sample, the results may represent something different than the broader population. The research focused solely on individual investors and did not consider the perspectives of institutional investors or other stakeholders in the SRI industry.

The study's practical implications are twofold. First, knowing how behavioral biases, such as herd behavior, overconfidence, and reactions to ESG news, affect SRI decisions can help investors and managers make better and more sustainable investment decisions. To reduce biases and encourage responsible investing, strategies might be created. In addition, the discovery of gender differences in SRI decisions, with women showing a stronger propensity, emphasizes the need for targeted marketing and communication strategies to promote more engagement in sustainable finance. These implications provide valuable insights for investors, managers, and policymakers seeking to advance sustainable investment practices.

The study has important social implications. It offers insights into the factors influencing individuals' SRI decisions, contributing to greater awareness and responsible investment practices. The gender disparities found in the study serve as a reminder of the importance of inclusivity in sustainable finance to promote balanced and equitable participation. Addressing these disparities can empower individuals of both genders to contribute to positive social and environmental change. Overall, the study encourages responsible investing and has a beneficial social impact by working towards a more sustainable and socially conscious financial system.

This study addresses a significant research gap by employing Bayesian linear regression method to examine the impact of behavioral biases on SRI decisions thereby offering more meaningful results compared to conventional frequentist estimation. Furthermore, the integration of behavioral finance with sustainable finance offers novel perspectives, contributing to the understanding of investors, investment managers, and policymakers, therefore, catalyzing responsible capital allocation. The study's exploration of gender dynamics adds a new dimension to the existing research on SRI and behavioral finance.