34254 research outputs found
Sort by
Types, Strategies, and Challenges of Organizations Promoting Time Banking in Taiwan
No full textTime banking is a reciprocal sharing system for exchange services. In this system, individuals invest their time in providing services or skills and earn time credits stored by an intermediary organization. These credits can then be used to exchange for services from others. In our country, various types of organizations currently utilize time banks, emphasizing the value of personal contribution and highlighting the feature of mutual assistance. However, existing literature lacks a comprehensive comparative analysis of the types of experiences in implementing time banks and an exploration of the difficulties and challenges encountered by organizations operating time banks. Therefore, this study adopts a qualitative research approach, conducting in-depth interviews with 14 time bank managers to gather diverse operational experiences. Secondary data are used for triangulation. Drawing on the CORPS model framework proposed by Situ Daxian (1999), this study categorizes the models of time banking in Taiwan and explores their coping strategies for challenges and difficulties.
Regarding research findings, this study first employs the indicators of "whether the time bank is embedded in existing organizations" and "whether membership qualifications are present," to propose three implementation types: member-based, task-based, and independent time banks. The study further analyzes the challenges facing organizations operating time banks, including the coexistence of volunteer service systems and time bank members leading to management conflicts, difficulties in developing service targets, insufficient human and financial resources, technological burdens, unclear service content, and a lack of performance indicators. Coping strategies include planning activities to attract service recipients, collaborating across departments to obtain human, material, and financial resources, using organizational design to facilitate service exchange, and innovating service content to transform difficulties.
Finally, this study has several practical suggestions. First, organizations promoting time banking could establish management systems for both time banks and volunteer services, cultivate talents to facilitate cross-departmental collaboration, and develop performance indicators for time banks. Second, the government needs to understand local needs, determine the institutional positioning of time banks, establish "exemplary time banks" for demand hotspots, adopt a "gradual change" approach within existing volunteer system institutions, and consistently engage in "information disclosure" to introduce the notion of time banking to the public
Exploring the Influence of Videoconference Fatigue on Presenteeism and Social Loafing
No full textSince the Covid-19 epidemic, many face-to-face meetings have been replaced by videoconference. Following this shift, there are plenty of reports and discussions about videoconference fatigue (also known as Zoom fatigue). In this study, we utilize TransactionBased odel of Stress (TBM) in order to discuss the impact of technostress which is induced by videoconference on stressors, situational factors, strains, and organizational outcomes of conference participants. This study aims to develop a comprehensive theoretical model which explores relationships between videoconference fatigue, presenteeism, and social loafing, while also examining the factors that influence videoconference fatigue.
We employed a quasi-experimental design combined with a questionnaire survey method and manipulated two important features of videoconference: the use of the camera and the use of the mute function. There are 204 participants who are students from a national university in the southern Taiwan. The results of the study indicated that videoconference fatigue indeed leads to behaviors that reduce team productivity, such as social loafing and presenteeism. Additionally, we also confirmed that the use of camera increases public self-awareness nd multitasking increases social anxiety, while both public self-awareness and multitasking effectively lead to the videoconference fatigue. These findings are expected to provide guidelines for remote teams to mitigate videoconference fatigue and offer suggestions to prevent behaviors that decrease productivity among conference participants
The impact of the passage of the new version of Act for the Development of Biotech and Pharmaceutical Industry on biotechnology industry companies
No full textDue to the impact of the COVID-19 epidemic on the world in recent years, the world has paid more attention to the importance of the development of the biotechnology industry. This study aims to understand \ue2the impact of the passage of the new version of Act for the Development of Biotech and Pharmaceutical Industry\ue2 and the tax incentive regulations on the stock prices of biotech companies, investment in new drug research, talent training, and equipment investment. The new regulations bring various favorable measures for biotech companies, and also raise problems about potential tax avoidance by enterprises. The revision also expands the scope of application for biotech and pharmaceutical companies, including companies involved in Contract Development and Manufacturing Organization (CDMO) for new drugs. The study uses event study analysis to categorize biotech companies into upstream, midstream, and downstream, and analyze the stock prices of companies involved in CDMO business to reflect the expectations of the Act on stock prices. Empirical results show that significant abnormal returns occurred after the event day, and the cumulative abnormal returns showed a significant positive relationship. This indicates that investors have an optimistic attitude towards this event. They are not particularly optimistic about the expansion of the scope of new drug companies due to the amendment of the regulation. Instead, they are more focused on the upstream companies that do not undertake commissioned development and manufacturing. It may be related to the distribution of industry sectors tailored to local conditions, with the upstream impact on the development of drugs and the manufacturing process of raw materials having a considerable degree of influence. To allow investors to more comprehensively assess the positive or negative impact of the new bill on the biotechnology industry and the situation caused by future trends in many aspects to provide relevant reference suggestions for the public\ue2s investment decision-making direction
Green learning based design of hybrid precoder in massive mimo systems
No full textThe advent of 5G technology has revolutionized communication networks, offering enhanced speed, increased data transfer capacity, low latency, and a reliable communication experience. In addressing the challenge of limited spectrum resources, particularly in high-frequency bands, millimeter-wave frequencies have emerged as a primary strategy for 5G communication. To harness the potential of millimeter-wave technology, the development of large-scale antenna systems has become imperative. However, traditional all-digital precoders pose challenges due to their high hardware costs. In response, hybrid precoders have been proposed as a cost-effective solution.
This paper explores the application of green learning, a logically structured training approach, in the design of hybrid precoders for large-scale antenna systems operating in the millimeter-wave frequency band. The goal is to reduce computational complexity while maintaining efficiency. The non-convex nature of power constraints for hybrid precoders poses a significant challenge, necessitating innovative approaches to mitigate computational costs and hardware requirements.
In contrast to prevailing methods that rely heavily on Singular Value Decomposition (SVD) based on channel matrix estimation, our proposed design approach leverages pilot-assisted transmission. Here, the received signal is utilized as the model input, eliminating the need for conventional channel estimation methods. This novel approach not only enhances computational efficiency but also streamlines hardware requirements.
Despite the prevailing trend in machine learning architectures to increase the number of neurons for performance improvement, this paper advocates for a green learning approach. By systematically analyzing data and selecting discriminative features, green learning enables model training with fewer data, achieving performance superior to other algorithms. Simulation results underscore the effectiveness of our proposed approach in balancing efficiency and computational complexity
Brain age prediction using 3D convolutional neural network on magnetic resonance imaging: comparison of non-deformed and deformed preprocessing
No full textThe study demonstrates the utilization of magnetic resonance brain imaging through convolutional neural network (CNN) to extract features of brain structure, successfully predicting the age of participants, termed as brain age prediction. Before inputting data into the CNN, a registration step aligns images from different databases and various participants to a common coordinate system, ensuring consistent presentation of the orientation and image coordinate directions of diverse brain images. Historically, non-linear registration methods were commonly used in the image preparation stage. However, this method distorts the brain structure, causing significant deformations in the images, and age-related features may be lost. Therefore, this study proposes a non-deforming image preprocessing pipeline, trained and predicted through CNN, and compared with the results of the deformation process.
The study collects 3D T1-weighted imaging (T1WI) from six publicly available databases. After excluding data with missing information, false images, or cases with brain disorders, a total of 3990 images from healthy participants are included, along with age, gender, and imaging main magnetic field strength, with an age range of 5-96 years. Image preprocessing steps include removing non-brain tissues, image registration, resampling, intensity normalization, brightness uniformity correction, and tissue segmentation. Two image registration methods are used: non-deforming and deforming registration, employing rigid and non-linear registration to align individual images to the standard brain template (MNI152 template), respectively. Images generated through non-deforming registration, including T1WI, gray matter, and white matter images, and images generated through deforming registration, including T1WI, gray matter, white matter, and Jacobian determinant, are separately used to train deep neural network models for age prediction.
The study uses a 3D CNN with a residual structure for training. Data are divided into nine age intervals, with 32 individuals selected as the validation set and 40 individuals as the test set for each interval. The remaining individuals constitute the training set. In each training epoch, a fixed number of training set data are chosen for each interval. The final model parameters are selected based on the epoch with the minimum validation set loss during the 1000 training epochs. Model performance is evaluated using mean absolute error (MAE), coefficient of determination (R2), and Pearson correlation coefficient (r).
Results show that both non-deforming and deforming models exhibit similar accuracy trends, with the T1 model outperforming single-tissue models (gray matter and white matter). However, in the deforming model, the Jacobian model performs worse than the other three models. Surprisingly, the non-deforming and deforming preprocessing methods show little difference in the prediction results of the T1 model. Unlike the expected results, preserving more original image information did not lead to an improvement in accuracy. One possible reason could be the significant differences in contrast among images from different participants in the non-deforming preprocessing method. After the non-deforming preprocessing method, although there is an improvement in the unevenness of the signal in each participant's image, there are still differences in the signal intensity and contrast between different participants' tissues. Integrating models with the same preprocessing method but different input images using average and linear regression resulted in higher accuracy compared to individual models. Lastly, additional testing using the NKI-RS database with 423 healthy participants indicates that the non-deforming T1, gray matter, and white matter models outperform the deforming models
Dynamic Trajectory Adjustment for UAV to Improve Channel Quality and Data Rate
No full textEmploying Unmanned Aerial Vehicle (UAV) shooting video in the air, the channel quality along its trajectory undergoes temporal changes, leading to fluctuations in the UAV's average Signal to Interference plus Noise Ratio (SINR) and throughput during flight, thereby affecting the quality of video playback. To address this challenge, we proposed a novel mechanism called Dynamic Trajectory Adjustment for UAV Channel Quality (DTA), which enhances channel quality by dynamically adjusting the UAV's trajectory during flight. The mechanism comprises two modules. In the first module, the signal range is divided into multiple discrete cubicles (CB) based on size. Given that Latitude, Longitude, and Altitude must be converted to 3D coordinates at ground level, we introduce a formula named Transformation of Latitude/ Longitude and 3D axis (TLLA). In the second module, we determine the received power of each CB with a central point. As each CB may contain noise and interference, calculating SINR requires accounting for their values. Subsequently, we convert the SINR of each CB to a Channel Quality Index (CQI) and compute the average CQI for all CBs. The proposed algorithm, Dynamic Adjustment of UAV Channel Quality (DACQ), compares the CQI of CBs surrounding the UAV. If CBs have the same CQI, we compare the interference values of surrounding CBs and dispatch the UAV to the one with the lowest value. To assess our proposed mechanism, we consider two interference levels (light and severe) and vary CB quantities. Simulation results show that increasing CB numbers under both interference levels notably enhances average SINR, as interference is confined to specific CBs. Additionally, we evaluate UAV channel quality (average SINR) and throughput across four trajectories, demonstrating that implementing the DACQ algorithm in UAV operations effectively enhances channel quality and throughput
A Political Economic Analysis on the Development of China\ue2s E-CNY/Digital Currency Electronic Payment (DCEP): A Structural Power Perspective
No full textThe purpose of this article is to explore why China is actively promoting the E-CNY policy as a research motivation. It delves into China\ue2s power demonstration and strategic influence by employing a structural power perspective.
In recent years, the global economy has witnessed a decoupling phenomenon in terms of security and trade, prompting reflection on the issues of national sovereignty erosion caused by globalization. Central bank digital currencies (CBDCs) have gradually become a strategic consideration for nations. As China\ue2s economic, military, and political power has strengthened, it has gained more leverage in international negotiations, with the E-CNY featuring prominently in China\ue2s \ue214th Five-Year Plan,\ue2 attracting attention from Western countries.
This article approaches the topic from the perspective of monetary power, with a focus on power discourse. Through the discussion of power dimensions proposed by Strange (security, production, finance, knowledge), it explores the strategic implications of China\ue2s push for the E-CNY. Emphasis is placed on its impact on domestic regulation and international currency competitiveness, ultimately contributing to the expansion of China\ue2s power
The Effect Of ISO14001, ISO14064, ISO14067 and ISO50001 On Operating Performance: The Case Of Cement, Steel, and Chemical industries
No full textThis research delves into the effects of International Organization for Standardization (ISO) environmental management system certifications on environmentally sensitive industries: cement, steel, and chemicals. Utilizing multiple regression analysis, the study evaluates the impact on business performance indicators (ROA, ROE, Tobin's Q) following ISO14001, ISO14064, ISO14067, and ISO50001 certifications.
The findings reveal that ISO certifications have a positive influence on enterprises, manifesting not only in resource efficiency and cost savings but also in operational, financial, and long-term market performance. This enhances a company's competitive stance. By focusing on resource management, risk control, and corporate image building, businesses can achieve a symbiotic relationship between environmental responsibility and economic gains. This study aims to comprehensively explore how corporate environmental management practices impact business performance, aiming to foster a harmonious coexistence of companies with the global environment
Portfolio Optimization Model with an Integrated Sliding Window Algorithm and Time Series Prediction Model
No full textDue to the rapid development of big data analytics, portfolio optimization with computational models has become a popular research field in business intelli- gence. Different methods to optimize portfolios have been developed, among them mean variance optimization and log-optimal methods are the most commonly used .
Different from the existing studies, this dissertation adopts log-optimal portfolio and integrates it with methods from other fields to devise an e\uef\uacicient op- timization model. Specifically, this study improves the sliding window approach by making its window size adjustable to serve as the investment period, and optimizing the model through hyper-parameter tuning with machine learning, in order to max- imize the return of investment. For evaluation, a signal decomposition method, i.e., Ensemble Empirical Mode Decomposition with Gated Recurrent Unit, is used, and the prediction values are leveraged for portfolio optimization. The results indicate the usefulness of the developed model to achieve a good performance in portfolio optimization, contributing toward research in big data analytics in business intelli- gence
The Effect of Stress and Positive/Negative Affect on Work Engagement : The Mediating Role of Emotional Exhaustion
No full textThe psychological work engagement of workers has always been the subject of hot research in the field of Human Resources Institute. In practice, this topic is a long-term problem that enterprises must continue to deal with. In this study, the job demand-resource model (JD-R model) is discussed, which is a popular work stress model in recent years, and it emphasizes that it can be applied to various workplaces. Most of the existing papers use European employees as samples to test hypotheses. This study uses Taiwanese employees as samples to clarify the role of emotional exhaustion in the model. Using daily diary questionnaires to collect data, a sample of 125 Taiwanese corporate employees was used to examine the relationship between perceived stress, positive and negative emotions, and work engagement, and to discuss the role of emotional exhaustion. The results of the regression analysis test support that negative emotions are negatively correlated with work engagement, and emotional exhaustion has a significant mediating effect; however, the relationship between positive and negative emotions and work engagement is not significant. These results show that the hypothesis of this model cannot be fully supported when tested with daily diaries, but also find the role and importance of emotional exhaustion in examining the degree of psychological well-being of employees. Future research could focus on whether the daily diary questionnaire made a difference