The goal of this thesis was to investigate the factors that influence the adoption of big data analytics by public sector employees based on the adapted Unified Theory of Acceptance and Use of Technology (UTAUT) model. A mixed method of survey and interviews were used to collect data from employees of a Canadian provincial government ministry. The results show that performance expectancy and facilitating conditions have significant positive effects on the adoption intention of big data analytics, while effort expectancy has a significant negative effect on the adoption intention of big data analytics. The result shows that social influence does not have a significant effect on adoption intention. In terms of moderating variables, the results show that gender moderates the effects of effort expectancy, social influence and facilitating condition; data experience moderates the effects of performance expectancy, effort expectancy and facilitating condition; and leadership moderates the effect of social influence. The moderation effects of age on performance expectancy, effort expectancy is significant for only employees in the 40 to 49 age group while the moderation effects of age on social influence is significant for employees that are 40 years and more. Based on the results, implications for public sector organizations planning to implement big data analytics were discussed and suggestions for further research were made. This research contributes to existing studies on the user adoption of big data analytics.

本研究主要先探討國內、外對外投資之相關文獻，再以經濟部統計處於2007年所實施的「製造業對外投資實況調查」問卷資料為分析對象，分別由廠商特性、產業特性及投資特性等3方面，運用probit model進行估計，來分析我國製造業廠商在從事對外投資活動後，對國內產品生產規模所產生的影響。經本研究實證發現，就廠商特性而言，「廠商規模」及「研發支出總額」為影響國內產品生產規模擴大的重要因素。在產業特性方面，則以「產業型態」及「對外投資地區」為重要影響因素，而若以投資特性來看，屬擴張型對外投資動機的「當地市場發展潛力大」、由台灣所提供之「原料進貨來源比率」及「零組件與半成品進貨來源比率」等因素為重要影響因素。 / After reviewing literature on outward foreign direct investment, this research conducts an empirical research based on 2007 statistical data from Ministry of Economic Affairs in Taiwan. We employ probit model to analyze the impacts of outward foreign direct investment on output of manufacturing industry in Taiwan. The empirical results show that the ‘firm size’ and ‘R&D expenditure’ categorized into firm characteristics, and ‘type of industry’ and ‘investment area’ classified into industry characteristics are statistically significant. Moreover, the expansionary FDI measured by ‘high potential of local market’, ‘rate of raw material purchased from Taiwan’ and ‘rate of components and semi-finished product purchased from Taiwan’ are also statistically significant.

Au Niger, les bonnes performances macroéconomiques enregistrées, consécutivement à la mise en œuvre du document stratégique de réduction de la pauvreté, suscitent la question de leur impact par rapport à l'évolution de l'inégalité et de la pauvreté. La présente recherche se propose d'analyser les spécificités des dynamiques d'inégalité et de pauvreté inhérentes au nouveau processus de développement, à partir de données d'enquêtes auprès des ménages entre 2005 et 2007/2008. Dans un premier temps, l'analyse, en statique comparative, indique une légère baisse des privations monétaires au Niger. Toutefois, des disparités semblent prévaloir entre les zones rurales et urbaines du pays. Globalement, la distribution des dépenses n'est pas inégalitaire, et le processus de croissance économique se révèle pro-pauvres au Niger, sauf dans la capitale où la croissance semble être pro-riches. Dans un second temps, la prise en compte de l'hétérogénéité de la pauvreté à travers la distinction entre la pauvreté chronique et transitoire, en relation avec la vulnérabilité, précise davantage l'appréhension des privations. D'une part, si la pauvreté chronique a sensiblement baissé, on note une progression de la pauvreté transitoire. D'autre part, l'étude souligne la forte vulnérabilité des ménages nigériens, notamment les ménages non pauvres qui ont une probabilité élevée d'exposition au risque de pauvreté à court terme. Finalement, l'analyse des privations au Niger est approfondie en intégrant une approche non monétaire. Les résultats obtenus confirment la complémentarité des approches mesurant les privations dans le cas du Niger.

Selection of signi?cant genes via expression patterns is important in a microarray problem. Owing to small sample size and large number of variables (genes), the selection process can be unstable. This research proposes a hierarchical Bayesian model for gene (variable) selection. We employ latent variables in a regression setting and use a Bayesian mixture prior to perform the variable selection. Due to the binary nature of the data, the posterior distributions of the parameters are not in explicit form, and we need to use a combination of truncated sampling and Markov Chain Monte Carlo (MCMC) based computation techniques to simulate the posterior distributions. The Bayesian model is ?exible enough to identify the signi?cant genes as well as to perform future predictions. The method is applied to cancer classi?cation via cDNA microarrays. In particular, the genes BRCA1 and BRCA2 are associated with a hereditary disposition to breast cancer, and the method is used to identify the set of signi?cant genes to classify BRCA1 and others. Microarray data can also be applied to survival models. We address the issue of how to reduce the dimension in building model by selecting signi?cant genes as well as assessing the estimated survival curves. Additionally, we consider the wellknown Weibull regression and semiparametric proportional hazards (PH) models for survival analysis. With microarray data, we need to consider the case where the number of covariates p exceeds the number of samples n. Speci?cally, for a given vector of response values, which are times to event (death or censored times) and p gene expressions (covariates), we address the issue of how to reduce the dimension by selecting the responsible genes, which are controlling the survival time. This approach enables us to estimate the survival curve when n << p. In our approach, rather than ?xing the number of selected genes, we will assign a prior distribution to this number. The approach creates additional ?exibility by allowing the imposition of constraints, such as bounding the dimension via a prior, which in e?ect works as a penalty. To implement our methodology, we use a Markov Chain Monte Carlo (MCMC) method. We demonstrate the use of the methodology with (a) di?use large B??cell lymphoma (DLBCL) complementary DNA (cDNA) data and (b) Breast Carcinoma data. Lastly, we propose a mixture of Dirichlet process models using discrete wavelet transform for a curve clustering. In order to characterize these time??course gene expresssions, we consider them as trajectory functions of time and gene??speci?c parameters and obtain their wavelet coe?cients by a discrete wavelet transform. We then build cluster curves using a mixture of Dirichlet process priors.

This dissertation consists of four papers, each with an application of a discrete dependent variable model, censored regression or duration model to a credit market phenomenon or monetary policy question. The first three essays deal with bank lending policy, while the last one studies interest rate policy by Central Banks. In the first essay, a bivariate probit model is estimated to contrast the factors that influence banks’ loan granting decision and individuals’ risk of default. This model is used as a tool to construct a Value at Risk measure of the credit risk involved in a portfolio of consumer loans and to investigate the efficiency of bank lending policy. The second essay takes the conclusions from the first paper as a starting point. It investigates if the fact that banks do not minimize default risk can be explained by the existence of return maximization policy. For this purpose, a Tobit model with sample selection effects and variable censoring limits is developed and estimated on the survival times of consumer loans. The third paper focuses on dormancy, instead of default risk or survival time, as the most important factor affecting risk and return in bank lending. By means of a duration model the factors determining the transition from an active status to dormancy are studied. The estimated model is used to predict the expected durations to dormancy and to analyze the expected profitability for a sample loan applicants. In the fourth paper, the discrete nature of Central Bank interest rate policy is studied. A grouped data model, that can take the long periods of time without changes in the repo rate by the Central Bank into account, is estimated on weekly Swedish data. The model is found to be reasonably good at predicting interest rate changes. / Diss. (sammanfattning) Stockholm : Handelshögsk.

Selection of signi?cant genes via expression patterns is important in a microarray problem. Owing to small sample size and large number of variables (genes), the selection process can be unstable. This research proposes a hierarchical Bayesian model for gene (variable) selection. We employ latent variables in a regression setting and use a Bayesian mixture prior to perform the variable selection. Due to the binary nature of the data, the posterior distributions of the parameters are not in explicit form, and we need to use a combination of truncated sampling and Markov Chain Monte Carlo (MCMC) based computation techniques to simulate the posterior distributions. The Bayesian model is ?exible enough to identify the signi?cant genes as well as to perform future predictions. The method is applied to cancer classi?cation via cDNA microarrays. In particular, the genes BRCA1 and BRCA2 are associated with a hereditary disposition to breast cancer, and the method is used to identify the set of signi?cant genes to classify BRCA1 and others. Microarray data can also be applied to survival models. We address the issue of how to reduce the dimension in building model by selecting signi?cant genes as well as assessing the estimated survival curves. Additionally, we consider the wellknown Weibull regression and semiparametric proportional hazards (PH) models for survival analysis. With microarray data, we need to consider the case where the number of covariates p exceeds the number of samples n. Speci?cally, for a given vector of response values, which are times to event (death or censored times) and p gene expressions (covariates), we address the issue of how to reduce the dimension by selecting the responsible genes, which are controlling the survival time. This approach enables us to estimate the survival curve when n << p. In our approach, rather than ?xing the number of selected genes, we will assign a prior distribution to this number. The approach creates additional ?exibility by allowing the imposition of constraints, such as bounding the dimension via a prior, which in e?ect works as a penalty. To implement our methodology, we use a Markov Chain Monte Carlo (MCMC) method. We demonstrate the use of the methodology with (a) di?use large B??cell lymphoma (DLBCL) complementary DNA (cDNA) data and (b) Breast Carcinoma data. Lastly, we propose a mixture of Dirichlet process models using discrete wavelet transform for a curve clustering. In order to characterize these time??course gene expresssions, we consider them as trajectory functions of time and gene??speci?c parameters and obtain their wavelet coe?cients by a discrete wavelet transform. We then build cluster curves using a mixture of Dirichlet process priors.

民國94年7月1日開始施行的勞工退休金條例，是政府近年來對勞工退休規劃的重要施政政策，依據此條例規定，員工超過二百人以上的企業，多了企業年金保險的選項可供選擇。而依據第14條第三款規定，勞工得在其每月工資百分之六範圍內，自願另行提繳退休金。勞工自願提繳部份，得自當年度個人綜合所得稅總額中全數扣除。本研究開始之時適逢勞退新制實施之初，並無企業年金中個人相對提撥之實際資料，因此藉分析個人年金保險保戶之屬性及對年金保險消費型態之探討，推論企業年金中選擇自願提繳之個人因素，供各界參考。 本研究整理與回顧國內外相關之職業退休金制度，並探討OECD各國企業年金的運作方式，以及國內年金市場的結構與產品。在實証模型上，則利用線性迴歸模型（OLS），分析影響年金保險保額的原因，並利用間斷性機率模型Probit Model 探討影響傳統型年金保險或投資型年金保險的因素。 / The Labor Pension Act, one of the major policies for the labor retirement planning of the R.O.C. government, was officially put into practice on July 1st, 2005. Based on the regulation of the Act, companies with more than 200 employees will have the Annuity Insurance as alternatives. According to the Article 14 -3 of the Act, a worker may voluntarily contribute per month, up to 6% of his/her monthly wages to his/her pension fund account. The full amount of the voluntary pension contribution made by a worker may be deducted from the worker's taxable income in the year concerned. Therefore, this research intends to analysis between the attributions and consumption behaviors of the employees joining in the policy of the Annuity Insurance, and then generalizes the factors why the workers choose the voluntary pension contribution policy. In this research, I would compare the pension policies used in different countries, look into the ways that the OECD are running their Enterprise Annuity policies, and evaluate the various pension policies. By using the real diagnosis Model, I would use the OLS to analysis the influences over Annuity Insurance Insured value and then use the Probit Model to explore the influences over traditional Annuity Insurance and the Investment Annuity Insurance.

Research on road safety has been of great interest to engineers and planners for decades. Regardless of modeling techniques, a serious factor of inaccuracy - in most past studies - has been data aggregation. Nowadays, most freeways are equipped with continuous surveillance systems making disaggregate traffic data readily available ; these have been used in few studies. In this context, the main objective of this dissertation is to capitalize highway traffic data collected on a real-time basis at the moment of accident occurrence in order to expand previous road safety work and to highlight potential further applications. To this end, we first examine the effects of various traffic parameters on type of road crash as well as on the injury level sustained by vehicle occupants involved in accidents, while controlling for environmental and geometric factors. Probit models are specified on 4-years of data from the A4-A86 highway section in the Ile-de -France region, France. Empirical findings indicate that crash type can almost exclusively be defined by the prevailing traffic conditions shortly before its occurrence. Increased traffic volume is found to have a consistently positive effect on severity, while speed has a differential effect on severity depending on flow conditions. We then establish a conceptual framework for incident management applications using real-time traffic data on urban freeways. We use dissertation previous findings to explore potential implications towards incident propensity detection and enhanced management

Using the E. coli vaccine as a case study, this thesis examines the factors affecting the adoption of technologies with positive spillover (externality) effects related to food safety. Positive spillovers occur when the benefits from a technological innovation extend beyond the firm (farm) adopting the technology or they do not flow to the adopter. If there are insufficient incentives for the firm to adopt the new technology, adoption levels are sub-optimal, resulting in forgone benefits to society. These benefits include the avoidance of potential health costs, productivity loss and premature death costs as a result to exposure to E. coli O157:H7. Therefore, if the market incentives to adopt the technology are strengthened, adoption levels of the technology could reach socially optimal levels resulting in an improvement in food safety. This has been the case in the Canadian cattle industry, where the uptake of the E. coli vaccine by cow-calf producers has been very low. As such, a number of potential incentives to increase adoption of the vaccine were identified and assessed through a survey of cow-calf producers on the Prairies. Data from the survey were analyzed using a stated preference methodology, Best-Worst Scaling, and Latent Class cluster analysis. A Binary Probit Model was also used to examine the factors affecting willingness to adopt the vaccine. The results suggest that a significant number of producers were not aware of the existence of the E. coli vaccine. In addition, producers were most likely to be influenced in their adoption decisions by market/supply chain oriented incentives and government intervention incentives in the form of subsidies. On the other hand, incentives that were least likely to influence cow-calf producers’ decisions to adopt included government intervention through recommending use of vaccine and neighbours (other cow-calf producers) adopting the vaccine. The Latent Class cluster analysis revealed the existence of three unique producer clusters with different attitudes towards these incentives. Several socio-demographic variables and individual characteristics utilized in the Probit analysis were found to be determinants of a producer’s willingness to adopt an E. coli vaccine. The implications of this research are such that producer education and awareness campaigns may be utilized as tools for disseminating information on food safety technologies such as the E. coli vaccine. Furthermore, the market/supply chain incentives may be used to form potential market-based solutions to address the current low adoption rates. The existence of three unique producer clusters suggest that a one-size fits all strategy to encourage the adoption of the E. coli vaccine might be difficult to implement and thus a more targeted approach may be a feasible alternative.

The credit scoring with the help of classification techniques provides to take easy and quick decisions in lending. However, no definite consensus has been reached with regard to the best method for credit scoring and in what conditions the methods performs best. Although a huge range of classification techniques has been used in this area, the logistic regression has been seen an important tool and used very widely in studies. This study aims to examine accuracy and bias properties in parameter estimation of the logistic regression by using Monte Carlo simulations in four aspect which are dimension of the sets, length, the included percentage defaults in data and effect of variables on estimation. Moreover, application of some important statistical and non-statistical methods on Turkish credit default data is provided and the method accuracies are compared for Turkish market. Finally, ratings on the results of best method is done by using receiver operating characteristic curve.