The quest towards resolving concerns about transportation energy consumption and emissions across nations has created more interests to investigate factors responsible for households’ car ownership. While literature holds an extensive body of investigation usually compartmentalised in individual different disciplines, limited efforts have been made to promote inter-linkages of this strand of research across different disciplines. To fill this gap, this study developed an integrating Multinomial logit (MNL) model to examine the impact of some rarely-investigated and conventional explanatory variables, including: ethnicity, accommodation tenure, settlement nature, mental belief, environmental concern, geographical regions, household structure, driving licence, number of household income earners and household income, on car ownership. Analysis based on rich data sets of British Household Survey and US Consumer Expenditure Survey found not only the conventional explanatory variables to be significantly linked to the number of cars in the US and UK households, but also the rarely-investigated psychological variables were found to be significantly linked as well. As Socio-demography, Geography and Psychology impact on how people and households process information and assess market offers (e.g., products and services), this study presents findings which have beneficial implications for policymakers and transportations planners, including those who would like to alter people’s behaviour from private car ownership to public transportation use, car sellers in terms of how to identify and reach potential customers, provision of alternative forecasting approaches to car ownership scholars as well as possible consideration for general car ownership decision making. Caution should be taken when interpreting the relationship between psychological factors and car ownership since the psychological factors adopted are measure representatives from databases used with limitations in the factor structure for a representative sample of the countries’ population.

In this paper, we first present the basic principles of set theory and combinatorial analysis which are the most useful tools in computing probabilities. Then, we show some important properties derived from axioms of probability. Conditional probabilities come into play not only when some partial information is available, but also as a tool to compute probabilities more easily, even when partial information is unavailable. Then, the concept of random variable and its some related properties are introduced. For univariate random variables, we introduce the basic properties of some common discrete and continuous distributions. The important properties of jointly distributed random variables are also considered. Some inequalities, the law of large numbers and the central limit theorem are discussed. Finally, we introduce additional topics the Poisson process.

Instance segmentation has a great potential for improving the current state of littering by autonomously detecting and segmenting different categories of litter. With this information, litter could, for example, be geotagged to aid litter pickers or to give precise locational information to unmanned vehicles for autonomous litter collection. Land-based litter instance segmentation is a relatively unexplored field, and this study aims to give a comparison of the instance segmentation models Mask R-CNN and DetectoRS using the multiclass litter dataset called Trash Annotations in Context (TACO) in conjunction with the Common Objects in Context precision and recall scores. TACO is an imbalanced dataset, and therefore imbalanced data-handling is addressed, exercising a second-order relation iterative stratified split, and additionally oversampling when training Mask R-CNN. Mask R-CNN without oversampling resulted in a segmentation of 0.127 mAP, and with oversampling 0.163 mAP. DetectoRS achieved 0.167 segmentation mAP, and improves the segmentation mAP of small objects most noticeably, with a factor of at least 2, which is important within the litter domain since small objects such as cigarettes are overrepresented. In contrast, oversampling with Mask R-CNN does not seem to improve the general precision of small and medium objects, but only improves the detection of large objects. It is concluded that DetectoRS improves results compared to Mask R-CNN, as well does oversampling. However, using a dataset that cannot have an all-class representation for train, validation, and test splits, together with an iterative stratification that does not guarantee all-class representations, makes it hard for future works to do exact comparisons to this study. Results are therefore approximate considering using all categories since 12 categories are missing from the test set, where 4 of those were impossible to split into train, validation, and test set. Further image collection and annotation to mitigate the imbalance would most noticeably improve results since results depend on class-averaged values. Doing oversampling with DetectoRS would also help improve results. There is also the option to combine the two datasets TACO and MJU-Waste to enforce training of more categories.

本研究著眼於台灣經濟轉型、中國經濟的崛起與台灣企業組織的變化，從台灣企業集團的總部設立、跨界投資的區位選擇及在地再投資三個面向進行討論，期望在既有的研究基礎上，就台灣廠商在兩岸投資區位佈點的考量提出完整性的觀察，並強化既有的研究。 首先，以台灣1000大製造業為研究樣本，選擇包括純辦公室使用、研發設計、台商一千大、跨國生產網絡、外資企業、員工人數、資本總額、知識密集型、傳統型製造業等變項分別代表總部功能、跨界治理能力及企業屬性三大類變數，透過二元羅吉特模型以及集群分析方法，探討台灣企業在首都、都會區以及生產性服務業及創新氛圍同質性地區的總部設立區位選擇行為。實證的結果發現，代表企業屬性變數的資產總額、員工人數和產業別明顯影響台灣製造業廠商在首都設立總部的區位選擇，而總部功能為純辦公室使用或設有研發機構者更傾向將總部設立於首都或都會區，跨界治理能力的影響則未能獲得證實。另外，過去國內在研究企業總部地點選擇研究上較少從創新氛圍角度出發，而本研究實證的結果發現，台灣製造業廠商企業總部的區位選擇不僅受到地區生產性服務業的影響，也受到地區創新氛圍的影響。 在跨界投資區位選擇部分，本研究以台灣250大企業集團中的知識密集型製造業集團為研究對象，以台灣、環渤海地區、長江三角洲地區、珠江三角洲地區為研究場域，選擇企業特性與投資區位條件變數，並以多項羅吉特模型進行實證分析。其中，企業特性變數為產業類別、投資經驗、投資時間等三項因子，而投資區位條件則有勞工薪資、市場規模、區域創新強度及外資投資強度等因子。實證結果發現，代表經濟發展階段的投資時間變項確實會影響企業集團的區位選擇行為，產業的類別不同其區位選擇也會不同，先前的投資經驗雖然影響區位選擇。但是與過去研究不同的是，本次實證發現對台灣企業來說面對相似而且鄰近的市場，進入新市場的動機可能比過去的投資經驗來得重要的多，同時投資區位條件亦會影響區位選擇行為。另外，過去較少直接連結廠商生產面的區域創新能力亦明顯影響企業集團的區位選擇，因此本研究認為區域創新活動對於跨國企業在地化取得知識及技術亦具有相當重要的意義。 在地再投資部分以台灣製造業1000大廠商中知識密集型製造業為研究對象，並以工業地域觀點所劃分的台灣地區北、中、南三大區域為研究場域，選擇包括在台投資經驗、總部區位、第一次投資決策、路徑依循等企業廠商組織決策之屬性變數，以及包含區域中科學園區的設立、產業專業化係數、雜異化指標等區域環境變數，透過多項羅吉特模型進行實證分析。實證的結果發現，總部區位確實影響後續再投資的工廠區位選擇，第一次的投資決策經驗對於第二次投資的區位選擇行為影響比總部區位的影響明顯，代表時間演進而產生路徑相依的地區經濟型態差異變項也確實會影響區位選擇行為。而當區域內科學園區的發展相較未臻成熟時，其區域的賦能仍不足以吸引企業廠商進駐，至於台灣企業的再投資區位選擇基於對區域特性的了解較偏好區域內工業地域的地方化經濟，而不偏好區域內工業地域的都市化經濟。 / Stressed on the Taiwanese economical transition, the up-rising of Chinese economy and the change of Taiwanese enterprise organization as well as based on the past research, this study explores the factors affecting location selection behavior of Taiwanese firms across Taiwan Strait from three aspects including the establishment of enterprise headquarter, cross-border investment and local re-investment. On the establishment of enterprise headquarter, the top 1000 manufacturing firms in Taiwan were sampled and some factors were analyzed including office type, R&D, multinational production network, foreign enterprise, number of employee, total asset, knowledge-intensive business, and traditional manufacturing firms. However, these factors could be classed into three fields: headquarter function, cross-border management ability and firm characteristics. Then, the location selection behavior of Taiwanese enterprise headquarter was examined by the techniques of binary logit model and cluster analysis technique among capital area, urban area and homogenous area with productive service industry and innovation-based cluster. The results of empirical analysis show that the factors represented firm characteristics including total asset, number of employee and enterprise type significantly affected the location selection of Taiwanese enterprise headquarter. Furthermore, it is also verified that the enterprise headquarter had been established in capital or urban area if the headquarter was provided with R&D or simply used as office, but the effect of cross-border management upon headquarter establishment is insignificant. The effect of innovation-based cluster upon location selection of enterprise headquarter is seldom studied in the past. However, according to empirical results in this study, they show that location selection of Taiwanese enterprise headquarter is affected not only by local Productive Service industry but also by regional innovation-based cluster. On the location selection of cross-border investment, this study focused on the area of Taiwan, Bohai Economic Rim, Yangtze River Delta and Pearl River Delta. The top 250 Taiwanese enterprise groups were taken into consideration, and the multinomial Logit model was adopted for empirical analysis in which firm characteristics and location conditions were chosen as research variables. Where, firm characteristics contained industrial type, investing experience and investment time, and location conditions included labor cost, market scale, regional innovation intensity and foreign investment intensity. The empirical results indicate that industrial type and investment time significantly affect the selection of investment locations. In contrast, investment experience only slightly influences the selection of investment locations. In addition, we find that entrepreneurial motivation to enter new markets may be much more influential than prior location investment experiences for Taiwanese enterprises functioning within similar markets. Regional differences shaping investment conditions in Taiwan and mainland China also affect the selection of investment locations. Our analysis shows a particularly strong linkage between regional innovation capacity and the selection of investment locations. This implies that regional innovation capacity plays a very important role in the selection of investment locations for multinational enterprises On local re-investment, the top 1000 knowledge-intensive manufacturers in Taiwan were the samples divided by region into the northern, central and southern Taiwan groups by administrative region. The factors affecting organizational decisions were the attribute variables, including Taiwan investment experience, headquarters location, first investment experience and path dependence; and the factors affecting location selection were the regional environment variables, including regional science park status, industry specialization coefficient and Hirschman-Herfindahl index (HHI). The multinomial Logit model was used for empirical analysis, and the results show that the headquarters location affects plant location selection in re-investment, and the first investment experience has a more significant effect on the plant location selection in the second investment than the headquarters location, suggesting that the path-dependent heterogeneity in regional economic style developed over time affects location selection. Also, the immaturity of regional science parks affects plant location selection when regional empowerment cannot attract enterprises. Lastly, Taiwanese enterprises prefer regions with localized economies to regions with urbanized economies for plant location selection.

Multivariate mixed Poisson processes are special multivariate counting processes whose coordinates are, in general, dependent. The first part of this thesis is devoted to properties which multivariate counting processes may possess. Such properties are, for example, the Markov property, the multinomial property and regularity. With regard to regularity we study the properties of transition probabilities and intensities. The second part of this thesis restricts the class of all multivariate counting processes by additional assumptions leading to different types of multivariate mixed Poisson processes which, however, are connected with each other. Using a multivariate version of the Bernstein-Widder theorem, it is shown that multivariate mixed Poisson processes are characterized by the multinomial property. Furthermore, regularity of multivariate mixed Poisson processes and properties of their moments are studied in detail. Throughout this thesis, two types of stability of properties of multivariate counting processes are studied: It is shown that most properties of a multivariate counting process are stable under certain linear transformations including the selection of single coordinates and summation of all coordinates. It is also shown that the different types of multivariate mixed Poisson processes under consideration are in a certain sense stable in time.

Saskatchewan is the only province in Canada to have routinely tested the antimicrobial susceptibility of all provincially reported human cases of campylobacteriosis. From 1999 to 2006, 1378 human Campylobacter species infections were tested for susceptibility at the Saskatchewan Disease Control Laboratory using the Canadian Integrated Program for Antimicrobial Resistance Surveillance panel and minimum inhibitory concentration (MIC) breakpoints. Of these, 1200 were C. jejuni, 129 were C. coli, with the remaining made up of C. lari, C. laridis, C. upsaliensis and undifferentiated Campylobacter species. Campylobacter coli had significantly higher prevalences of ciprofloxacin resistance (CIPr), erythromycin resistance (ERYr), combined CIPr-ERYr resistance and multidrug resistance (to three or greater drug classes) than C. jejuni. Logistic regression models indicated that CIPr in C. jejuni decreased from 1999 to 2004 and subsequently increased in 2005 and 2006. The risk of CIPr was significantly increased in the winter months (January to March) compared to other seasons. A comparison of logistic regression and Cox proportional hazard survival models found that the latter were better able to detect significant temporal trends in CIPr and tetracycline resistance by directly modeling MICs, but that these trends were more difficult to interpret. Scan statistics detected significant spatial clusters of CIPr C. jejuni infections in urban centers (Saskatoon and Regina) and temporal clusters in the winter months; the space-time permutation model did not detect any space-time clusters. Bernoulli scan tests were computationally the fastest for cluster detection, compared to ordinal MIC and multinomial antibiogram models. eBURST analysis of antibiogram patterns showed a marked distinction between case and non-case isolates from the scan statistic clusters. Multilevel logistic regression models detected significant individual and regional contextual risk factors for infection with CIPr C. jejuni. Patients infected in the winter, that were between the ages of 40-45 years of age, that lived in urban regions and that lived in regions of moderately high poultry density had higher risks of a resistant infection. These results advance the epidemiologic knowledge of CIPr C. jejuni in Saskatchewan and provide novel analytical methods for antimicrobial resistance surveillance data in Canada. / Saskatchewan Disease Control Laboratory (Saskatchewan Ministry of Health); Laboratory for Foodborne Zoonoses (Public Health Agency of Canada); Centre for Foodborne, Environmental and Zoonotic Infectious Diseases (Public Health Agency of Canada); Ontario Veterinary College Blake Graham Fellowship

Existing research suggests that when consumers encounter hybrid products or boundary-spanning products with attributes belonging to multiple categories, consumers tend to generate inferences based on only a single product category. Reliance on a single category for inferencing is termed as the “single category belief problem” which has been regarded as a vital marketing challenge because it leads consumers to underestimate the true utility of a hybrid product as certain product attributes are ignored. Our objective was to explore whether single category beliefs manifest in consumer choice for a hybrid product when strategically placed within varying contexts. The research used discrete choice experiment (DCE) to test hypotheses. Our research confirms that the single category belief is evident in consumer choice. We also found that the context the hybrid product is placed within has a major influence on what consumers preferred the most. Depending on the context a hybrid product was seen in had significant influence on how consumers evaluated product attributes and made purchase decisions. The findings for this research may be very beneficial for marketers.

近年來，國內有許多研究提倡大眾運輸導向發展 (Transit Oriented Development)的理念，政府也大力推動大眾運輸系統的建設，其中最為重要的是捷運的建設，在台北都會區，捷運路網的建設正逐步完成，與捷運建設息息相關的捷運聯合開發(Transit Jointed Development)也隨之蓬勃發展，同時，捷運聯合開發亦是我國推動大眾運輸導向發展普遍的作法之一。目前台北都會區目前共有82處聯合開發基地，已完工基地有35處，可容納6,317個家戶，以及755,773.69帄方公尺樓地板面積，對於減緩日益嚴重的都市住宅問題，有一定程度的幫助。但在規劃聯合開發住宅時，聯合開發住宅在不同類型、特性之捷運場站中，將面臨到許多問題，如：純住宅型態或住商混合型態較能符合民眾之需求？哪些家戶會選擇聯合開發住宅？又其家戶類型以及選擇原因為何？這些問題如何解決，係本研究欲探討之內容，因此，本研究以台北捷運聯合開發已完工且辦理租售作業之開發基地作為研究對象，並篩選出9處聯合開發基地進行實證研究，透過問卷調查的方式，瞭解民眾之聯合開發住宅選擇行為與旅運行為，並透過二項與多項羅吉特模型，探討影響民眾聯合開發住宅選擇行為之影響因素。 實證結果發現，在旅運行為方面，聯合開發住宅住戶之大眾運輸使用率大幅增加，在通勤時間與花費方面，通勤時間與花費均減少。除此之外，聯合開發住宅住戶之汽車持有率與使用頻率均大幅減少。在家戶特性方面，捷運聯合開發住宅住戶之家戶規模普遍較小，且家計負責人之年紀普遍較為年輕，進一步形成其他特性，如：就學人口比例較低、家戶月收入較低等。在影響因素方面，家戶規模、住宅帄均單價、住宅規模對民眾選擇不同類型之聯合開發住宅有顯著影響。最後依據實證結果，建議未來聯合開發住宅之規劃應加入TOD的規劃原則，對於聯合開發住宅之坪數、商業面積，應依捷運場站之類型進行調整，使聯合開發住宅之效益達到最大。 / In recent years, there are many researches promote the idea of the transit-oriented development. The government also vigorously promotes this infrastructure projects. One of the most important projects is the development of the MRT system. In Taipei metropolitan area, while the construction of the MRT network is gradually completed, the transit jointed dvelopment is also flourishing. Moreover, transit jointed development is the most common way in order to promote TOD. Currently, there are 82 Transit Jointed Development bases in Taipei metropolitan area. 35 of the bases have already completed, which can accommodate 6,317 household with 75,577,369 square meters of floor area. It certainly will help to alleviate the problem of urban housing. However, in planning of the Jointed Residential Development, it will face many problems due to the different types of characteristics of the MRT station. For example, which households will choose a jointed development dewilling? What is the reason of choosing jointed development dewilling? How to solve these problems? These are the contents of the study. Therefore, in this study, we target the bases that have already been completed and applied for rental operations in transit jointed development as the research object, and select 9 of them for the empirical research. In order to understand people’s choice behavior in jointed development dewilling, we use survey as a method, and explore the factors that affect people’s choice behavior by applying Binary Logit and Multinomial Logit Models. The results of empirical research show that households in the jointed development dewilling increase their public transport usage and reduce their commuting time and costs. In addition, their car ownership rate and frequency of use are significantly reduced. In the aspect of household characteristics, the households in the jointed development in the household scale are generally small and relatively young age of the householder. Those characteristics are going to further the formation of the other features, such as: the lower the proportion of student population and lower income households. Impact factors, such as the size of the household, the average residential price, residential scale on the people choose different types of joint development dewilling have a significant effect. Finally, according to the empirical results, it is recommended that the TOD should be added to the planning principle of the future Joint Residential Development. For the Jointed Residential Development, the size of house and commercial area should be adjusted along with the MRT in order to maximize the efficiency.

Cette recherche se penche sur l’impact des facteurs contextuels, familiaux et individuels dans l’accessibilité au postsecondaire des Premières Nations et des Métis âgés entre 18 et 34 ans qui vivent hors collectivités des Premières Nations au Canada. En plus de l’impact de ces facteurs au niveau national, cette recherche s’intéresse aussi à leur impact selon les trois types de systèmes scolaires canadiens « à choix exclusif », « à choix progressif » et « à choix multiples ». Des régressions logistiques multinomiales ont été réalisées sur les données de l’enquête auprès des peuples Autochtones de 2012. Trois groupes de facteurs qui totalisent 24 indicateurs ont été examinés afin de répondre aux questions de recherche. Les résultats de cette recherche montrent que même si les facteurs contextuels, tels que l’environnement social et scolaire au secondaire et les facteurs familiaux, tels que la scolarité des membres de la famille sont déterminants dans l’accès au postsecondaire des Autochtones, les facteurs individuels sont les principaux facteurs prédictifs de cet accès. Parmi les facteurs les plus importants, il y a la performance académique et le décrochage scolaire, suivi de l’âge qui affecte considérablement cet accès, particulièrement à l’université. De plus, l’analyse de l’effet de modération selon les trois systèmes scolaires canadiens a également permis de distinguer une différence significative entre ces derniers qui est perceptible, d’une part, par les variances expliquées et, d’autre part, par l’impact différé des trois groupes de facteurs étudiés. / This research assesses the impact of contextual, family and individual factors in access to postsecondary education of First Nations and Métis people aged between 18 and 34 years old, living outside First Nations communities in Canada. This research examines also the impact of these factors across the three postsecondary types of school systems of Canada, i.e., “exclusive choice”, “progressive choice” and “multiple choice”. Multinomial logistic regression analyses were performed on the data of the Aboriginal Peoples Survey of 2012. Three groups of factors totaling 24 indicators were examined in order to answer the research questions. The results of this research show that, while contextual factors, such as the social and academic environment and family factors, such as the high academic level of family’s members are determinant in access to postsecondary education among Aboriginals living outside First Nations communities, the individual factors are the main predictors of this access. Among these main predictors, there are academic performance, dropping out of school and age. In addition, the analysis of moderation effects according to the three postsecondary types of systems present in Canada showed significant differences that can be seen, first, in the explained variance and second, in the difference between the importance of the different groups of factors in each system.

Multivariate mixed Poisson processes are special multivariate counting processes whose coordinates are, in general, dependent. The first part of this thesis is devoted to properties which multivariate counting processes may possess. Such properties are, for example, the Markov property, the multinomial property and regularity. With regard to regularity we study the properties of transition probabilities and intensities. The second part of this thesis restricts the class of all multivariate counting processes by additional assumptions leading to different types of multivariate mixed Poisson processes which, however, are connected with each other. Using a multivariate version of the Bernstein-Widder theorem, it is shown that multivariate mixed Poisson processes are characterized by the multinomial property. Furthermore, regularity of multivariate mixed Poisson processes and properties of their moments are studied in detail. Throughout this thesis, two types of stability of properties of multivariate counting processes are studied: It is shown that most properties of a multivariate counting process are stable under certain linear transformations including the selection of single coordinates and summation of all coordinates. It is also shown that the different types of multivariate mixed Poisson processes under consideration are in a certain sense stable in time.