Search

This thesis presents a history of sports for individuals with physical disabilities in Taiwan between 1945 and 2007. A Foucauldian lens is adopted. Primary data was collected from personal archives of key informants who had lived experience with the developments during this period and with whom interviews were also conducted. Materials were organised using Foucauldian genealogy and analysis was conducting using the lens of biopower to investigate how the bodies of people with physical disabilities were organised in relation to power in the rehabilitation and sporting context. The research portrays how, between 1945 and 1971, the epidemics of poliomyelitis left a large number of children in Taiwan paralysed, and most of them were segregated in institutions. These children became a social problem, and in these institutions, rehabilitation was regarded as physical education and thus became the hallmark of the special education at this time. Development in sports for the disabled occurred between 1972 to 1992, as the authorities began to pay attention to sports for these individuals and it was employed as a tool to encourage them to contribute to the greater national good. Between 1993 and 2007, the status of individuals with physical disabilities entered a liminal state in which they were neither fully included nor excluded in the sports for the disabled in Taiwan. The thesis concludes by highlighting how the Taiwanese State's approach to individuals with physical disabilities targeted them as a social problem that needed to be managed, and sports were employed as techniques to fulfil this goal. The disabled body then becomes a pawn in the State's power game which transformed over time, and it was made docile through the techniques of normalisation that are central to engagement with rehabilitation, physical education and sport.

Human decision-making is subject to a number of cognitive biases. One of the first biases to be identified is the way information in a decision problem is framed. It is now widely recognized that when information is framed positively (i.e. in terms of gains), people make risk-averse choices, whereas when information is framed negatively (i.e. in terms of losses), people make risk-seeking choices. Information framing is one of the most studied cognitive biases and its effects have been observed in decision problems ranging from medicine and finance to terrorism. More recently, researchers have focused on the role of emotion in the decision-making process. The purpose of this research was to investigate the role of emotion on the effects of information framing. To do this, I developed a technology-related information-framing problem about the reliability of internet service providers. Having observed strong information framing effects in my technology-related problem, I used Lerner et al.'s (2015) Emotion Imbued Choice (EIC) model to investigate the effects of integral and incidental emotions. Contrary to the predictions of the EIC, I found that the characteristics of options (in this case, positively framed or negatively framed information) did not affect integral emotions. In order to investigate the effects of incidental emotions, I developed a method to induce participants with fear and happiness by using the International Affective Picture System (IAPS). Consistent with the predictions of the EIC, I found that incidental emotions can moderate the effects of information framing (e.g.higher levels of incidental fear can lead participants to make risk-averse choices when information is framed negatively, even though negatively framed information usually leads to risk-seeking choices). The main theoretical contribution in this thesis comes from an amended EIC model that takes into consideration the effects of information framing. Other researchers could also use the method I used to induce emotions using the IAPS. This method was found to induce emotions more reliably than some existing methods, and as it does not involve language production, can be used with mixed-cultural and cross-cultural samples. There are also several practical implications of my research, which I consider in this thesis.

The relationships between service quality, consumer satisfaction, and loyalty have been widely discussed in the service marketing literature, but there is still an ongoing debate about these relationships because they have not been well studied in spectator sport industry (Theodorakis & Alexandris, 2008). Schmitt (2011) claimed that consumer experiential could be a new perspective for evaluating the nature of these relationships. However, experience marketing has some special and unique attributes requiring specific definitions in specific research contexts. Thus, the current research attempts to (1) identify what kinds of experiences are found in Taiwanese professional baseball games; (2) test models of relationships between consumer experience, service quality, satisfaction and loyalty taking into account the unique aspects of the specific context of Taiwanese professional baseball games; and (3) test the role of consumer experience on the relationships between perceived service quality, satisfaction and loyalty. In order to achieve the research aims, the current research adopted a mixed method approach using both qualitative and quantitative methods. Four focus groups, representing the qualitative stage, were conducted to acquire a better understanding of consumer experience types and characteristics in Taiwanese professional baseball games. The main findings of the focus groups are that (1) ordinary experiences include the game itself while additional events and activities characterise extraordinary experiences. However, (2) when games involve a special performance and record or modifies a ranking position, they can be seen as extraordinary experiences as well. Moreover, ordinary experience and extraordinary experiences also depend on clubs, game days, and stadiums. A questionnaire survey, representing the quantitative stage, was used to investigate the relationships between consume experience, service quality, consumer satisfaction and loyalty. The sampling strategy was designed based on the specific research context features (i.e., different clubs, dates and stadia), and 1,229 questionnaires were collected. A confirmatory factor analysis was conducted in order to test the measurement model and compare theoretical models according to specific research context features. The main findings were that: (1) the degree of association between consumer experience and service quality is significant, (2) both direct and indirect relationships of consumer experience on consumer satisfaction and loyalty are significant, (3) only an indirect influence, via consumer satisfaction, of service quality on consumer loyalty is significant, (4) the differences of relationships are significant between clubs, but not significant between dates and stadia. The current research concluded that (1) consumer experience is an important dimension to take into consideration to better understand the formation of consumer loyalty, (2) there is a two-way relationship between service quality and consumer experience, (3) Intangible experience elements are captured by consumer experience while tangible experience elements are captured by service quality, (4) Extraordinary experiences in sport spectating events are highly dependent on situational factors, (5) specific research contexts features are important to take into consideration when conducting consumer experience research.

The advancement of machine learning algorithms promoted its adoption in a wide variety of applications, from recommending content in online systems to evaluating risks of defendants for judges. As these algorithms are increasingly being used for important applications, it is vitally important to make these algorithms transparent and fair to all stakeholders affected by these algorithms, specifically for algorithms that inform or make consequential decisions for people. Many current machine learning explanations and visualizations systems are designed for data scientists and practitioners, which might not be interpretable for the real stakeholders. At the same time, many algorithms also do not take into account the fairness notions and viewpoints of the stakeholders. The research in this dissertation focuses on promoting the explainability and fairness for these algorithms, by involving relevant stakeholders in the process. For explainability, we investigate the effectiveness of different explanation strategies that communicate how a university admission algorithm works for non-expert stakeholders. To promote fairness, we propose a framework for eliciting stakeholders’ subjective fairness notions of a child maltreatment predictive system.Lastly, we investigate in real world, how involving stakeholders in a decision-making algorithm impacts the disproportionate decisions between racial groups. These work brings greater understanding on how to create machine learning systems that respect the values of human stakeholders to benefit society.

Hurricane Florence in 2018 was one of the worst in North Carolina history. The catastrophic flooding and infrastructure damage endangered millions of residents and visitors. The focus of this paper is the development of methods to understand evacuation behavior to aid in the development of future evacuation plans and emergency responses. While existing hurricane evacuation studies primarily use statistical models based on survey data, we develop a new approach using location data passively collected by cellphones. By analyzing location information from about 18,000 anonymous and opted-in users before and during the hurricane, we discern their home census area and whether they evacuated, and if they evacuated, where they evacuated too. About 91% of inferred home areas are found to be located on residential land parcels. North Carolina has 100 counties. Of those 100 counties, about 21 counties were in areas where the condition warranted the consideration of evacuation. In those 21 counites the evacuation rate was about 56%. In about six coastal counties, conditions were most severe and for those six counties, the evacuation rate was found to be about 67%, which is consistent with survey data collected from 1,421 respondents.

Enzymes are biological catalysts which greatly accelerate the rate of chemical reactions with remarkable substrate specificity and stereoselectivity. To optimize their catalytic abilities, many enzymes have also evolved to cooperate with coenzymes. The variety of coenzymes largely enhances the scope of enzyme-catalyzed reactions that are not accessible to the twenty canonical amino acids. In general, the chiral enzyme active sites allow enzymes to selectively bind substrates with a specific conformation which can lower the activation energy of the reactions by stabilizing high energy intermediates. Therefore, if the substrate of an enzyme contains chirality, in many cases, enzymes preferably accept one of the stereoisomers for catalysis. Inspired by this phenomenon, the work described in this dissertation focuses on mechanistic studies of two enzyme-catalyzed unusual chemical reactions by using chemically synthesized chiral substrate analogues. The first section of this dissertation focuses on 1-aminocyclopropane-1-carboxylic acid deaminase (ACCD), an enzyme that plays a role in regulating the production of the potent plant hormone, ethylene. ACCD is a pyridoxal-5ʹ-phosphate (PLP)-dependent enzyme that catalyzes a C-C bond cleavage event that is unique among the PLP-dependent enzymes. (R)- and (S)-2,2-difluoro-1-aminocyclopropane-1-carboxylic acid are synthesized as mechanism based inhibitors. Our studies suggest that the previously proposed acid-catalyzed mechanism for the ACCD-catalyzed reaction is less likely, and instead, a nucleophilic attack mechanism is consistent with the accumulated experimental results from the mechanistic studies of the ACCD reaction. The second section of this dissertation focuses on (E)-4-hydroxyl-3-methylbutenyl-1-diphosphate reductase (IspH), an essential enzyme in the non-mevalonate pathway (MEP pathway) for isoprenoid biosynthesis that employs a [4Fe-4S] cluster as cofactor for catalysis. Stereospecific labeling of tritium on the substrate and substrate analogue of IspH are used to generate chiral methyl groups in the products of the IspH reaction via catalysis in D2O-based buffer condition. Our results suggest that a configuration inversion at C4 of the substrate during the catalysis of IspH reaction is a universal step among all currently proposed mechanisms for the IspH reaction. Furthermore, we also provide the first experimental evidence for the direction of protonation of the allyl anion intermediate during IspH catalysis.

Abstract: Due to its wide application in various fields, clustering, as a fundamental unsupervised learning problem, has been intensively investigated over the past few decades. Unfortunately, standard clustering formulations are known to be computationally intractable. Although many convex relaxations of clustering have been proposed to overcome the challenge of computational intractability, current formulations of clustering remain largely restricted to spherical Gaussian or discriminative models and are susceptible to imbalanced clusters. To address these shortcomings, we propose a new class of convex relaxations that can be flexibly applied to more general forms of Bregman divergence clustering. By basing these new formulations on normalized equivalence relation matrix, we retain additional control on relaxation quality, which allows improvement in clustering quality. We fur- thermore develop optimization methods that improve scalability by exploiting recent implicit matrix norm methods. We find that the new formulations are able to efficiently produce tighter clusterings that improve the accuracy of state of the art methods.

This thesis investigates the capacity expansion in the interaction between the operational and financial contexts. Estimating demand before a firm expands its capacity is a key activity in operations management, but it is difficult and unlikely to be observed by researchers and practitioners when the market is imperfectly competitive. This thesis uses the index share to identify demand, realising the simultaneity of matching demand with supply in the empirical capacity expansion. Moreover, the operating and financial performance effects of capacity expansion are also worth investigating, as these reflect the firm's profitability and financial reaction associated with capacity expansion. In this thesis, firm profit, stock return, and firm value are applied to measure the performance outcomes, and the influence of index share is also considered to build the interaction of capacity expansion and financial implication. In addition, as a significant expenditure in the process of capacity expansion, particularly in the semiconductor manufacturing sector, fixed cost plays a critical role in determining capacity policy; however, only few studies have investigated this due to the discontinuity and nonlinearity derived from the consideration of fixed cost. An economic structure is employed to figure out this issue and compare the models with and without fixed cost based on different measuring criteria. Therefore, the interactive relationships of capacity expansion decision with demand, firm performance, and fixed cost are well investigated in the thesis. To evaluate models, an empirical case of the semiconductor manufacturing sector is used for each topic, where data are obtained from US-listed semiconductor manufacturing firms in the iShares PHLX semiconductor ETF (SOX) from 2006 to 2010. Drawing on findings and outcomes specified in the empirical estimates and counterfactual analyses, many managerial implications are proposed; for example, the debtholder may prefer to invest in firms with small index share as they facilitate capacity expansion through fully utilising financial funding. This thesis contributes to the operations-finance interface in many ways. In general, the studies of interaction between operational and financial decisions, such as capacity expansion and debt, are advanced and extended through building systematic frameworks of simultaneous determinates. Moreover, with the use of Bayesian estimation, the theoretical models are empirically examined; for instance, the important role of fixed cost on capacity expansion and supply chain coordination is verified by the case of the semiconductor manufacturing sector. Besides, the methodology proposed in this thesis can be extended to more general forms and is applicable in the empirical analyses for other sectors, including the automobile and chemical industries.

The main focus of my research is to design effective learning techniques for information retrieval and mining in high-dimensional databases. There are two main aspects in the retrieval and mining research: accuracy and efficiency. The accuracy problem is how to return results which can better match the ground truth, and the efficiency problem is how to evaluate users' requests and execute learning algorithms as fast as possible. However, these problems are non-trivial because of the complexity of the high-level semantic concepts, the heterogeneous natures of the feature space, the high dimensionality of data representations and the size of the databases. My dissertation is dedicated to addressing these issues. Specifically, my work has five main contributions as follows. The first contribution is a novel manifold learning algorithm, Local and Global Structures Preserving Projection (LGSPP), which defines salient low-dimensional representations for the high-dimensional data. A small number of projection directions are sought in order to properly preserve the local and global structures for the original data. Specifically, two groups of points are extracted for each individual point in the dataset: the first group contains the nearest neighbors of the point, and the other set are a few sampled points far away from the point. These two point sets respectively characterize the local and global structures with regard to the data point. The objective of the embedding is to minimize the distances of the points in each local neighborhood and also to disperse the points far away from their respective remote points in the original space. In this way, the relationships between the data in the original space are well preserved with little distortions. The second contribution is a new constrained clustering algorithm. Conventionally, clustering is an unsupervised learning problem, which systematically partitions a dataset into a small set of clusters such that data in each cluster appear similar to each other compared with those in other clusters. In the proposal, the partial human knowledge is exploited to find better clustering results. Two kinds of constraints are integrated into the clustering algorithm. One is the must-link constraint, indicating that the involved two points belong to the same cluster. On the other hand, the cannot-link constraint denotes that two points are not within the same cluster. Given the input constraints, data points are arranged into small groups and a graph is constructed to preserve the semantic relations between these groups. The assignment procedure makes a best effort to assign each group to a feasible cluster without violating the constraints. The theoretical analysis reveals that the probability of data points being assigned to the true clusters is much higher by the new proposal, compared to conventional methods. In general, the new scheme can produce clusters which can better match the ground truth and respect the semantic relations between points inferred from the constraints. The third contribution is a unified framework for partition-based dimension reduction techniques, which allows efficient similarity retrieval in the high-dimensional data space. Recent similarity search techniques, such as Piecewise Aggregate Approximation (PAA), Segmented Means (SMEAN) and Mean-Standard deviation (MS), prove to be very effective in reducing data dimensionality by partitioning dimensions into subsets and extracting aggregate values from each dimension subset. These partition-based techniques have many advantages including very efficient multi-phased pruning while being simple to implement. They, however, are not adaptive to different characteristics of data in diverse applications. In this study, a unified framework for these partition-based techniques is proposed and the issue of dimension partitions is examined in this framework. An investigation of the relationships of query selectivity and the dimension partition schemes discovers indicators which can predict the performance of a partitioning setting. Accordingly, a greedy algorithm is designed to effectively determine a good partitioning of data dimensions so that the performance of the reduction technique is robust with regard to different datasets. The fourth contribution is an effective similarity search technique in the database of point sets. In the conventional model, an object corresponds to a single vector. In the proposed study, an object is represented by a set of points. In general, this new representation can be used in many real-world applications and carries much more local information, but the retrieval and learning problems become very challenging. The Hausdorff distance is the common distance function to measure the similarity between two point sets, however, this metric is sensitive to outliers in the data. To address this issue, a novel similarity function is defined to better capture the proximity of two objects, in which a one-to-one mapping is established between vectors of the two objects. The optimal mapping minimizes the sum of distances between each paired points. The overall distance of the optimal matching is robust and has high retrieval accuracy. The computation of the new distance function is formulated into the classical assignment problem. The lower-bounding techniques and early-stop mechanism are also proposed to significantly accelerate the expensive similarity search process. The classification problem over the point-set data is called Multiple Instance Learning (MIL) in the machine learning community in which a vector is an instance and an object is a bag of instances. The fifth contribution is to convert the MIL problem into a standard supervised learning in the conventional vector space. Specially, feature vectors of bags are grouped into clusters. Each object is then denoted as a bag of cluster labels, and common patterns of each category are discovered, each of which is further reconstructed into a bag of features. Accordingly, a bag is effectively mapped into a feature space defined by the distances from this bag to all the derived patterns. The standard supervised learning algorithms can be applied to classify objects into pre-defined categories. The results demonstrate that the proposal has better classification accuracy compared to other state-of-the-art techniques. In the future, I will continue to explore my research in large-scale data analysis algorithms, applications and system developments. Especially, I am interested in applications to analyze the massive volume of online data.

In modern geometric modeling, one of the requirements for interpolants is that they 'preserve shape well.' Shape preservation has often been associated with preservation of monotonicity and convexity/concavity. While shape preservation cannot yet be defined quantitatively, it is generally agreed that shape preservation involves eliminating extraneous non-physical oscillation. Classical splines, which exhibit extraneous oscillation, do not 'preserve shape well.' Recently, Lavery introduced a new class of cubic L1 splines. Empirical experiment has shown that cubic L1 splines are cable of providing C¹-smooth, shape-preserving, multi-scale interpolation of arbitrary data, including data with abrupt changes in spacing and magnitude, with no need for monotonicity or convexity constraints, node adjustment or other user input. However, the shape-preserving capability of cubic L1 splines has not been proved theoretically. The currently available algorithm only provides an approximation to the coefficients of cubic L1 splines. To lay the groundwork for theoretical analysis and the development of an exact algorithm, this dissertation proposes to treat cubic L1 spline problems in a geometric programming framework. Such a framework leads to a geometric dual problem with a linear objective function and convex quadratic constraints. It also provides a linear system for dual-to-primal conversion. We prove that cubic L1 splines preserve shape well, in particular, in eliminating non-physical oscillations, without review of raw data or any human intervention. We also show that cubic L1 splines perform well for multi-scale data, as well as preserve linearity and convexity/concavity under mild conditions. An exact algorithm based on the geometric programming model is proposed for solving cubic L1 splines. It decomposes the geometric programming problem into several independent small-sized sub-problems and applies a specialized active set algorithm to solve the sub-problems. The algorithm is numerically stable and highly parallelizable. It requires only simple algebraic operations.