34254 research outputs found
Sort by
Composition and sources of particulate organic matter in the northeastern Taiwan Strait in summer
No full textSuspended particulate matter (SPM) serves as the carrier in material transport processes, with marine SPM being a crucial node for terrestrial material transport from land to sea. The biogeochemical composition of SPM helps to identify the physicochemical processes occurring during the transport. The Taiwan Strait, connecting the East China Sea and the South China Sea, is a vital conduit for material exchange in the East Asian marginal seas. The dissolved substances and surface sediments in the Taiwan Strait have been studied to a considerable extent, but the geochemical composition of the SPM remains insufficiently studied. This study focuses on the northeastern segment of the Taiwan Strait, primarily influenced by mountainous rivers in northwestern Taiwan. By comparing SPM samples collected from both seawater and rivers in summer, this study aims to investigate the fate of particulate organic matter in the northern Taiwan Strait from its source to the sink.
The results showed that the marine organic matter in the study area was generally fresh, the freshwater water-affecting coastal region had higher concentrations of marine organic matter, and chlorophyll a correlated well with marine organic carbon estimated by the stable isotopic values of organic carbon (\uce\ub413COC). The concentration of lignin, a biomarker of higher terrestrial plant, had a poor correlation with salinity and a fair correlation with SPM in the study area, suggesting resuspension to be the primary controlling factor for the seawater lignin content. Lignin concentration exhibited a poor correlation with \uce\ub413COC-estimated terrestrial organic carbon. This study further compared lignin data from riverine SPM, marine SPM, and marine sediment. The results showed that the ratio of lignin to sediment supply in the catchment was relatively stable, with mild dilution of lignin caused by the large amount of sediment generated during typhoons. The lignin content in marine SPM was much lower than that in river SPM possibly caused by lignin degradation and the contribution from low-lignin suspended sediment. There was only a moderate correlation in the lignin stock between the water column and surface sediment, suggesting that the spatial distribution of sedimentary lignin is affected not only by vertical settling or resuspension, but also lateral transport
Convergence Analysis for Discrete-Time Model Coming From Social Network
No full textThis thesis presents a novel framework for a general discrete-time nonlinear opin-
ion dynamics model driven by signed graphs in cooperative-antagonistic social networks
(Chapter 3). The framework demonstrates potential applications across various mod-
els, including the general discrete-time nonlinear polar opinion model and the Degroot-
Friedkin model. Moreover, it is mentioned that the framework even can be applied to the
analysis of disconnected graphs. When it comes to the application of the framework to a
general discrete-time nonlinear polar opinion model (Chapter 4), we provide a comprehen-
sive stability analysis that compasses both unsigned and signed directed graphs. For un-
signed graphs, consensus is achieved in the general discrete-time nonlinear polar opinion
model. Additionally, a detailed characterization of system stability under signed directed
graphs with external stationary points is provided. The results presented under weaker
graph conditions and precise steady-state values are characterized. Finally, leveraging
the framework, the consensus problem in unsigned directed graphs within the Degroot-
Friedkin model is analyzed (Chapter 5). Unlike previous approaches, we rather employ
skillful transformations, including reduced and error systems to simplify the proof signif-
icantly
Decentralized Offloading Strategy for Non-Arbitrarily Divisible Tasks
No full textIn recent years, with the rapid rise of the Internet of Things (IoT) and mobile networks, numerous products and applications have emerged, bringing significant improvements to our life. However, while this technology has provided convenience, other challenges have also arisen that need to be overcome. Firstly, the rapid increase in network demand has led to congestion in the backbone network, severely impacting the overall network quality. Secondly, emerging applications require lower network transmission latency but the existing network architecture cannot provide this service.
To address these issues, the European Telecommunication Standards Institute (ETSI) proposed Mobile Edge Computing (MEC) in 2014. MEC is a novel network architecture that provides computational capabilities at the edge of the mobile network. Users can directly obtain the required data through computation devices deployed on base stations, eliminating the need to wait for responses from cloud servers. This reduces service latency and decreases the overall network load.
This thesis is based on the environment of Mobile Edge Computing, exploring how user-generated tasks can be offloaded to edge servers and scheduled appropriately within the edge computing environment. The goal is to ensure tasks can be completed within their constraints and efficiently utilize resources in the edge computing environment. While numerous publications in this field have adopted a centralized architecture for task dispatching, this approach faces challenges to scale effectively and may encounter single-point failures. Additionally, many papers assume tasks are either indivisible or can be arbitrarily divided, while in reality, tasks often have predetermined ways of segmentation instead of arbitrary segmentation. Therefore, the method proposed in this thesis adopts a non-centralized architecture to determine how tasks with segmentation limitations can be offloaded and scheduled
Utilizing Brain MRI for Alzheimer's Disease Classification and Longitudinal Analysis of Logical Recall Scores
No full textAlzheimer\ue2s disease is a neurodegenerative disorder characterized by the loss of cognitive memory and behavioral abilities. Due to the global aging population, the number of Alzheimer\ue2s disease patients is rapidly increasing, making it one of the significant epidemic diseases in many countries. Through classification and longitudinal analysis, we aim to identify the risk factors associated with disease recognition. And this study focused on structural brain magnetic resonance imaging (sMRI) axis, utilizing three-dimensional image information for analysis.
In terms of baseline classification, this research employs the ROI-based method and utilizes the three-dimensional brain segmentation software AssemblyNet for skull stripping and correction on brain images. The images are then divided into multiple sub-region volume estimates, considered as candidates for classification features. And we hope to find relevant information and analyze binary, ternary, and quaternary classification models using features beyond Clinical Dementia Rating (CDR) and Mini-Mental Examination Scale (MMSE). These features may include volume, cognition, blood characteristics, and genetics (APOE) to aid doctors in diagnosis and prediction.
In view of the fact that Alzheimer\ue2s disease is often accompanied by the loss of various behavioral abilities, among which cognitive memory stands out as one of the main symptoms. This article focuses on two scores, immediate logical memory, and delayed logical memory, as the research objectives. Initially, a linear regression model is employed to identify baseline scores and baseline volumes. Subsequently, we apply the adjustment of covariates, including demographic characteristics, clinical features, and disease group, to identify brain characteristics as impactful factors. Two linear mixed-effects models are then utilized: one for longitudinal scores and baseline volume, and the other for longitudinal scores and longitudinal volume. These models take into account fixed effects such as
demographic characteristics, time, clinical features, disease group, etc. Additionally, they consider random effects, including slope and intercept estimates generated by individual cases over time, to explore the correlation between logical memory scores and various brain regions. The objective is to identify brain characteristics that correlate with memory for disease prevention and control, aiming to prevent deterioration
A Study of the Cultural Policies Which Influence Taiwanese Opera
No full text\ue3\ue3As an important cultural asset of our country, Taiwanese Opera is facing some challenges, such as the loss of audience. It is worth noticing how to reserve and pass on Taiwanese Opera to future generation. The government is a major force in supporting Taiwanese Opera. As for whether the cultural policies of the public sector can effectively help tackle the problems Taiwanese Opera is encountering, the researcher uses policy evaluation to answer the question. In this research, first, the researcher reviews the history of Taiwanese Opera to realize why the art is facing difficulties. Besides, the researcher introduces some cultural policies related to Taiwanese Opera and public policy evaluation theories. To examine how the programs and projects derived from the policies have influenced Taiwanese Opera and how they can improve the difficulties encountered by Taiwanese Opera troupes, the researcher uses the methods of policy impact evaluation to analyze the programs and projects, including \ue2Taiwan Top Annual Grant Program for Performing Arts Groups,\ue2 \ue2Project for Spreading Branches and Leaves of Traditional Theatrical Troupes & Performance and Promotion Project for Successors of the Traditional Arts\ue2 and \ue2Arts Fun Online Stage Project,\ue2 which are served as the cases in the research. Three policy evaluation criteria are used, \ue2Performance,\ue2 \ue2Adequacy\ue2 and \ue2Responsiveness.\ue2 \ue2Performance\ue2 is used to examine whether a policy reaches its goals. Research questions are as follows: First, what are the goals of the programs and projects which are derived from cultural policies? And how do they influence Taiwanese Opera and solve the difficulties faced by Taiwanese Opera? Second, how do the programs and projects reach to the criteria of policy impact evaluation? And do they respond to the needs of their target population? How can they be improved? To discover what the effects of the programs and projects are and measure whether the cases meet the three policy evaluation criteria, the researcher interviewed the interested parties, collecting related documents and making observations on policy outcomes.
\ue3\ue3Related to research findings, the outcome of policy impact evaluation is presented. For the first case \ue2Taiwan Top Annual Grant Program for Performing Arts Groups,\ue2 it meets \ue2Performance\ue2 and \ue2Responsiveness,\ue2 bettering the quality of creation and the administrative capacity of the Taiwanese troupes who apply for the program. Nevertheless, it partly meets \ue2Adequacy,\ue2 for the elected troupes are often fixed and the program cannot help those who are not chosen. \ue2Project for Spreading Branches and Leaves of Traditional Theatrical Troupes\ue2 can assist troupes to make new theater productions, but there is room for improvement for the creation, which makes the project partly meet \ue2Performance.\ue2 Since the cost of a new production is high, unable to be performed at a temple and improve the quality of outdoor Taiwanese Opera, the project doesn\ue2t meet \ue2Adequacy.\ue2 As for \ue2Responsiveness,\ue2 the project partly meets the criterion, for the interviewees feel that some of their needs are not fulfilled by the project. Next, \ue2Performance and Promotion Project for Successors of the Traditional Arts.\ue2 partly meets the three criteria. Although the project helps enhance the performing skills of the applicants, it cannot necessarily let them become the successor of a troupe. Aiming to promote the development of online performance and attract new audiences through the new mode, \ue2Arts Fun Online Stage Project\ue2 partly meets \ue2Performance,\ue2 though. The project does help some Taiwanese Opera troupes to produce creative online performances during the Covid-19 pandemic period, but most of the troupes have a box-office failure. Accordingly, the project meets \ue2Adequacy\ue2 a little and doesn\ue2t meet \ue2Responsiveness.\ue2 In the conclusion, the researcher proposes some advice about how to better the programs and projects. For example, \ue2Taiwan Top Annual Grant Program for Performing Arts Groups\ue2 should enhance its subsidy budget, helping more small troupes. \ue2Project for Spreading Branches and Leaves of Traditional Theatrical Troupes\ue2 should put more emphasis on the revitalization of outdoor Taiwaness Opera
Marine Vibrio vulnificus Sepsis: Molecular Pathogenesis and Therapeutic strategies
No full textVibrio vulnificus, a causal halophilic gram-negative zoonotic bacterium from the sea, causes serious wound infections and septicemia. Once developing early sepsis, high inflammatory immune response leads to poor prognosis of patients. Studies have shown that Vibrio vulnificus infection can stimulate the human immune system to produce a large amounts of cytokines, which further activate platelets and vascular endothelial cells, leading to endothelial cell damage and abnormal coagulation. In recent years, there is increasing evidence for the involvement of purinergic P2X7 receptor and NLR family pyrin domain containing 3 protein (NLRP3). The aim of this study is to investigate the underlying pathogenic mechanisms and to explore novel therapies that may protect against V. vulnificus cytotoxicity.
The anti-pyroptotic effects of P2X7 receptor (P2X7R) antagonists were assessed in an experimental murine V. vulnificus infection. In vitro, the effects of P2X7R antagonists on V. vulnificus-induced anti-phagocytosis and pyroptosis in murine macrophages were examined.
Here, we report that infection of mouse macrophages with V.\ue2vulnificus triggers anti-phagocytic effects and pyroptotic inflammation via ATP-mediated purinergic P2X7R signaling. V.\ue2vulnificus promoted P2X7-dependent NF-\uce\ubaB p65 translocation, modulating the expression of the inflammasome sensor NLRP3, apoptosis-associated speck-like protein (ASC), and pyroptotic protein gasdermin D (GSDMD) in mouse macrophages. V.\ue2vulnificus induced the NLRP3/caspase-1 inflammasome signaling complex expression that drives GSDMD transmembrane pore formation and secretion of IL-1\uce\ub2, IL-18, and MIP-2. This effect was blocked by P2X7R antagonists, indicating that the P2X7R mediates GSDMD-related pyroptotic inflammation in macrophages through the NF-\uce\ubaB/NLRP3/caspase-1 signaling pathway. Furthermore, blockade of P2X7R reduced V. vulnificus-colony-forming units in the spleen, immune cell infiltration into the skin and lung tissues, and serum concentrations of IL-1\uce\ub2, IL-18, and MIP-2 in mice.
These results suggest that the P2X7 and NLRP3 pathways play a vital role in mediating phagocytosis by macrophages and pyroptotic inflammation during V.\ue2vulnificus infection , and these studies offer new opportunities for therapeutic intervention in such bacterial infections. We have investigated the conditioned medium of induced pluripotent stem cells and found that it could improve the prognosis of sepsis induced by traditional E. coli. In the future, we hope to apply this finding to sepsis caused by marine Vibrio vulnificus
A study on the relationship between elementary school students\ue2 perceived teacher leadership style and students\ue2 learning motivation
No full textThe aim of this study is to investigate the relationship between elementary school students' perception of teacher leadership styles and their learning motivation.The research adopted a questionnaire survey method and used the "Leadership Style Scale" and "Learning Motivation Scale" as research tools. The survey was conducted among fifth and sixth grade students from two public elementary schools in Kaohsiung City. A total of 330 questionnaires were distributed, with 310 returned, and 251 were deemed valid, resulting in an effective response rate of 81%. Based on the data analysis, the following conclusions were drawn:
(1)There are significant differences in the degree to which elementary school children of different grades perceive teachers\ue2 use of the laissez-faire leadership style.
(2)Higher levels of perception among elementary school students regarding teachers\ue2 use of the laissez-faire leadership style correspond to lower learning motivation.
(3)Higher levels of perception among elementary school students regarding teachers\ue2 use of the directive leadership style correspond to increased learning motivation.
(4)Higher levels of perception among elementary school students regarding teachers\ue2 use of the caring leadership style correspond to increased learning motivation.
(5)Higher levels of perception among elementary school students regarding teachers\ue2 use of the transformational leadership style correspond to increased learning motivation.
(6)Elementary school students\ue2 perception of teachers\ue2 use of the transformational leadership style is highly correlated with students\ue2 learning motivation.
(7)Elementary school students\ue2 perception of teachers\ue2 use of the directive leadership style is moderately correlated with students\ue2 learning motivation.
(8)Elementary school students\ue2 perception of teachers\ue2 use of the caring leadership style is highly correlated with students\ue2 learning motivation.
(9)Elementary school students\ue2 perception of teachers\ue2 use of the laissez-faire leadership style is negatively correlated with students\ue2 learning motivation.
Based on the research results, practical teaching suggestions are put forward for teachers in class leadership, which will serve as a reference for teachers' future teaching and research.
\ue
Simultaneous Face Detection and 6DoF Pose Estimation
No full textThis thesis introduces a multi-task prediction model capable of simultaneously performing face detection, facial keypoint localization, and 6DoF (six-degree-of-freedom) head pose estimation. We observe that most existing models can handle either face detection or head pose estimation as individual tasks, requiring different methods to obtain comprehensive facial information. Therefore, we aim to integrate these functionalities to generate comprehensive prediction results while addressing the limitations of existing models.
We propose a One-Stage model based on the lightweight MobileNetV2 network, characterized by ease of training. This model can predict the face position, facial keypoints, and head pose for each detected face. To perform head pose estimation, we need to generate six degrees of freedom head pose labels and design a loss function for the model to learn, ensuring that the predicted head pose labels closely approximate the ground truth.In the common evaluation test sets of WIDER FACE and FDDB for face detection, our model outperforms other lightweight face detection models. In the common evaluation test sets of AFLW2000-3D and BIWI for head pose estimation, our model's performance excels in certain metrics compared to other head pose estimation models
Privacy Affordance and Privacy Disclosure in Online Health Community: The Moderating Effect of Gender
No full textNowadays, in the age of internet development, the advancement of information technology, there are many social media platforms, and health information is also emphasized, and health information is also shared on social media. The amount of health information has always been limited, and the users have privacy concerns due to the influence of privacy protection, and the sharing of health information on social media is also limited because of this. Therefore, this study will explore the privacy availability, understand the relationship between privacy availability and privacy control beliefs, and then further adjust and check how to implicitly protect privacy, so that after these processes, users' privacy can be protected, and then the privacy disclosure can be enhanced, so that the health information on health community platforms can be further enriched.
This study adopts the SOR theory and combines the relevant aspects of privacy availability and privacy protection. A total of 321 valid questionnaires were collected through Google and Surveycake platforms, and data statistics and analysis were conducted. The collected data samples were examined for common method bias and statistics by SPSS V26 software, and further analyzed for validity, reliability, hypothesis testing, and path coefficients by using SmartPLS V4.0.9.9 version, and the related results were discussed.
The results of this study are as follows, 1. Privacy availability increases privacy protection and has a greater effect on males than females. 2.Privacy availability increases privacy control beliefs and has a greater effect on females than males.3. Privacy control beliefs increase privacy protection adjustments on platforms and have a greater effect on females than males.4. Privacy protection increases an individual user's willingness to disclose privacy and has a greater effect on females than males.5. Privacy control beliefs increase an individual user's willingness to disclose privacy and have a greater effect on males than males.6. Privacy availability will be affected by the repressive effect of privacy control beliefs and will not increase users' willingness to disclose, but will increase users' willingness to disclose if privacy control beliefs are not taken into account, and the effect is greater for women than for men
The Interaction of Brokerage Reports Between Good and Bad Periods
No full textThis study divides the market into periods of prosperity and recession using proxy variables, namely the VIX index representing the market perspective and investor sentiment representing retail investors' perspective, to measure the state of the stock market. It investigates whether Taiwanese securities firms\ue2 investment recommendations and earnings forecasts exhibit preferences and biases in both good and bad periods. By combining market and company characteristics, it explores the impact of various brokerage reports on individual stocks and the market as a whole. Previous research has found that brokerage reports indirectly convey more company information to investors, generating abnormal returns. Moreover, brokerage reports' quality, quantity, and behavior vary across different time periods and contexts. This study reveals that the Taiwanese stock market is influenced by brokerage reports, resulting in abnormal returns, with variations in the accuracy of brokerage revenue forecasts during different periods. Furthermore, the positive impact of brokerage rating changes on the market outweighs the negative impact, especially in the electronics industry