Search results

For privacy protection, federated learning based on data separation allows machine learning models to be trained on remote devices or in isolated data devices. However, due to the limited resources such as bandwidth and power of local devices, communication in federated learning can be much slower than in local computing. This study aims to improve communication efficiency by reducing the number of communication rounds and the size of information transmitted in each round.

This paper allows each user node to perform multiple local trainings, then upload the local model parameters to a central server. The central server updates the global model parameters by weighted averaging the parameter information. Based on this aggregation, user nodes first cluster the parameter information to be uploaded and then replace each value with the mean value of its cluster. Considering the asymmetry of the federated learning framework, adaptively select the optimal number of clusters required to compress the model information.

While maintaining the loss convergence rate similar to that of federated averaging, the test accuracy did not decrease significantly.

By compressing uplink traffic, the work can improve communication efficiency on dynamic networks with limited resources.

The COVID-19 pandemic has caused severe challenges to ethnic minorities in the UK. While the experiences of migrants are both complex and varied depending on individuals' social class, race, cultural proximity to the host country and acculturation levels, more in-depth studies are necessary to fully understand how COVID-19 affects specific migrant groups and their health. Taiwanese migrants were selected because they are an understudied group. Also, there were widespread differences in pandemic management between the UK and Taiwan, making this group an ideal case for understanding how their acculturation journey can be disrupted by a crisis.

Qualitative data were collected at two different time points, at the start of the UK pandemic (March/April 2020) and six months on (October/November 2020), to explore migrant coping experiences over time. Theoretically, the authors apply acculturation theory through the lens of coping, while discussing health-consumption practices, as empirical evidence.

Before the outbreak of the pandemic, participants worked hard to achieve high levels of integration in the UK. The pandemic changed this; participants faced unexpected changes in the UK’s sociocultural structures. They were forced to exercise the layered and complex “coping with coping” in a hostile host environment that signalled their new marginalised status. They faced impossible choices, from catching a life-threatening disease to being seen as overly cautious. Such experience, over time, challenged their integration to the host country, resulting in a loss of faith in the UK’s health system, consequently increasing separation from the host culture and society.

It is important to note that the Taiwanese sample recruited through Facebook community groups is biased and has a high level of homogeneity. These participants were well-integrated, middle-class migrants who were highly educated, relatively resourceful and active on social media. More studies are needed to fully understand the impact on well-being and acculturation of migrants from different cultural, contextual and social backgrounds. This being the case, the authors can speculate that migrants with less resource are likely to have found the pandemic experience even more challenging. More studies are needed to fully understand migrant experience from different backgrounds.

Public health policymakers are advised to dedicate more resources to understand migrants' experiences in the host country. In particular, this paper has shown how separation, especially if embraced temporarily, is not necessarily a negative outcome to be corrected with specific policies. It can be strategically adopted by migrants as a way of defending their health and well-being from an increasingly hostile environment. Migrants' home country experience provides vicarious learning opportunities to acquire good practices. Their voices should be encouraged rather than in favour of a surprising orthodox and rather singular approach in the discussion of public health management.

The paper has clear public health policy implications. Firstly, public health policymakers are advised to dedicate more resources to understand migrants' experiences in the host country. Acknowledging migrants' voice is a critical first step to contribute to the development of a fair and inclusive society. Secondly, to retain skilful migrants and avoid a future brain-drain, policymakers are advised to advance existing infrastructure to provide more incentives to support and retain migrant talents in the post-pandemic recovery phase.

This paper reveals how a group of previously well-integrated migrants had to exercise “coping with coping” during the COVID crisis. This experience, over time, challenged their integration to the host country, resulting in a loss of faith in the UK’s health system, consequently increasing separation from the host culture and society. It contributes to the understanding of acculturation by showing how a such crisis can significantly disrupt migrants' acculturation journey, challenging them to re-acculturate and reconsider their identity stance. It shows how separation was indeed a good option for migrants for protecting their well-being from a newly hostile host environment.

With the rapid growth of the domestic lending industry, assessing whether the borrower of each loan is at risk of default is a pressing issue for financial institutions. Although there are some models that can handle such problems well, there are still some shortcomings in some aspects. The purpose of this paper is to improve the accuracy of credit assessment models.

In this paper, three different stages are used to improve the classification performance of LSTM, so that financial institutions can more accurately identify borrowers at risk of default. The first approach is to use the K-Means-SMOTE algorithm to eliminate the imbalance within the class. In the second step, ResNet is used for feature extraction, and then two-layer LSTM is used for learning to strengthen the ability of neural networks to mine and utilize deep information. Finally, the model performance is improved by using the IDWPSO algorithm for optimization when debugging the neural network.

On two unbalanced datasets (category ratios of 700:1 and 3:1 respectively), the multi-stage improved model was compared with ten other models using accuracy, precision, specificity, recall, G-measure, F-measure and the nonparametric Wilcoxon test. It was demonstrated that the multi-stage improved model showed a more significant advantage in evaluating the imbalanced credit dataset.

In this paper, the parameters of the ResNet-LSTM hybrid neural network, which can fully mine and utilize the deep information, are tuned by an innovative intelligent optimization algorithm to strengthen the classification performance of the model.

Project-based organizations (PBOs) face challenges to enhance employee work engagement because of dynamic and constant role configuration. Accordingly, this study aims to explore how ethical leadership enhances employee work engagement from a sensemaking perspective.

This study used a questionnaire-based quantitative research design to collect data from 194 full-time employees in PBOs. The data were analyzed via partial least squares-structural equation modeling (PLS-SEM) technique to test hypotheses.

The findings show a positive relationship between ethical leadership and work engagement. Additionally, the relationship between ethical leadership and work engagement is mediated by two sensemaking mechanisms, i.e. goal commitment and prosocial.

This study deepens the understanding of how ethical leadership enhances work engagement in PBOs by providing two sensemaking mechanisms. By exploring the sensemaking process through which ethical leaders help employees construct identity, the findings contribute to the current literature on how ethical leadership enhances work engagement in PBOs.

This paper aims to propose a novel control scheme and offer a control parameter optimizer to achieve better automatic carrier landing. Carrier landing is a challenging work because of the severe sea conditions, high demand for accuracy and non-linearity and maneuvering coupling of the aircraft. Consequently, the automatic carrier landing system raises the need for a control scheme that combines high robustness, rapidity and accuracy. In addition, to exploit the capability of the proposed control scheme and alleviate the difficulty of manual parameter tuning, a control parameter optimizer is constructed.

A novel reference model is constructed by considering the desired state and the actual state as constrained generalized relative motion, which works as a virtual terminal spring-damper system. An improved particle swarm optimization algorithm with dynamic boundary adjustment and Pareto set analysis is introduced to optimize the control parameters.

The control parameter optimizer makes it efficient and effective to obtain well-tuned control parameters. Furthermore, the proposed control scheme with the optimized parameters can achieve safe carrier landings under various severe sea conditions.

The proposed control scheme shows stronger robustness, accuracy and rapidity than sliding-mode control and Proportion-integration-differentiation (PID). Also, the small number and efficiency of control parameters make this paper realize the first simultaneous optimization of all control parameters in the field of flight control.

The authors investigate the varying impact of three categories of conflicting consumer reviews (i.e. conflicting opinions on attributes of a product item, conflicting ratings of an item and the intensity of conflicting reviews of an item) on the potential customers' perceived informativeness, which is expected to affect the perceived correct purchase.

To test their proposed hypotheses, the authors conducted an experiment using a 2 × 2 × 2 factorial design for each conflict type comprising two levels (low vs high).

The results of this study found that conflicting opinions on product attributes can enhance potential customers' perceptions of informativeness and subsequent correct purchase decisions while conflicting ratings and the intensity of conflicting reviews can diminish potential customers' perceptions of informativeness. In addition, conflicting ratings negatively moderate the effect of conflicting attributes on perceived informativeness such that the positive effect of conflicting attributes on perceived informativeness will be less prominent when conflicting ratings are present (vs absent).

While potential customers are browsing product descriptions, reviews and comments from other purchasers are also playing a role in influencing a potential customer's purchase decision. However, given the different experiences and temperaments of individuals, the subjective remarks and ratings of individuals are sometimes inconsistent or even conflicting, which can lead to confusion among potential customers. The authors categorize the positive or negative effects of the three conflicting reviews based on the two dimensions of ease of capture and product diagnosticity. The findings can help platforms optimize the display of product reviews to help potential customers make more accurate purchase decisions.

We attempt to construct a grey three-way conflict analysis model with constraints to deal with correlated conflict problems with uncertain information.

In order to address these correlated conflict problems with uncertain information, considering the interactive influence and mutual restraints among agents and portraying their attitudes toward the conflict issues, we utilize grey numbers and three-way decisions to propose a grey three-way conflict analysis model with constraints. Firstly, based on the collected information, we introduced grey theory, calculated the degree of conflict between agents and then analyzed the conflict alliance based on the three-way decision theory. Finally, we designed a feedback mechanism to identify key agents and key conflict issues. A case verifies the effectiveness and practicability of the proposed model.

The results show that the proposed model can portray their attitudes toward conflict issues and effectively extract conflict-related information.

By employing this approach, we can provide the answers to Deja’s fundamental questions regarding Pawlak’s conflict analysis: “what are the underlying causes of conflict?” and “how can a viable consensus strategy be identified?”

This paper aims to empirically investigate the impact of the carbon intensity constraint policy (CICP) on green innovation.

This study takes the implementation of the CICP as a quasi-natural experiment and uses a quasi–difference-in-difference method to investigate the impact of the CICP on firm green innovation from a microeconomic perspective.

The CICP significantly limits the quality of firms’ green innovation. Among the range of green patents, the CICP distorts only patents related to CO₂ emissions. The inhibitory effect is more pronounced in non-state-owned enterprises and heavily polluting firms. R&D investment and green investor are identified as the main mechanism.

These findings provide evidence for the influence of the CICP on firm green innovation, which can guide policymakers in China and other emerging economies that prioritize carbon intensity constraint targets and the improvement of relevant auxiliary measures.

Governments and firms should have a comprehensive understanding of environmental policies and corporate behavior and need to mitigate the negative impact through a combination of measures.

This study contributes to the literature by providing additional empirical evidence regarding the two opposing sides of the ongoing debate on the positive or negative effects of CICP. It also provides new evidence on the policy effect of the CICP on firm green innovation, together with its mechanisms and heterogeneous influences.

This research examined the impacts of diversified income from trading caterpillar fungus on pastoral households' livestock production and income. The specific objectives were to identify the main factors underlying participation in caterpillar fungus trade and to explore the impacts of a diversified income from trading fungus on livestock production activities and income.

Data were collected from a pastoral household survey (n = 503) in five Tibetan Autonomous Prefectures. The authors employed propensity score matching (PSM) procedures to estimate the effects of participation in trading caterpillar fungus.

Pastoral households participating in caterpillar fungus activities maintain smaller herds, sell fewer animals for profit, slaughter more livestock for family consumption and experience fewer livestock deaths compared to nonparticipants. There is also some evidence that pastoral households participating in caterpillar fungus activities have a higher annual income compared to nonparticipants.

A direct measure of grassland degradation was not included due to the data limitation. The estimated average treatment effects could differ under different observed households' characteristics.

This study fills a gap in the literature on the impacts of diversified income on livestock production activities. The authors provide a new perspective on the controversy over the extraction of caterpillar fungus. This study contributes to exploring the dual role of income diversification in addressing poverty and grassland resource degradation for Tibetan pastoral communities.