The work in this thesis is concerned with the development and extension of techniques for the assessment of influence diagnostics in data that include censored observations. Various regression models with censored data are presented and we concentrate on two models which are the accelerated failure time model, where the errors are generated by mixtures of normal distributions,and the Cox proportional hazards model. For the former, both finite discrete and continuous mixtures are considered, and an EM algorithm is used to determine measures of influence for each case. For the Cox proportional hazards model, various approaches to approximating influence curves are investigated. One-step or few-step approximations are developed using an EM algorithm and compared with a Newton-Raphson approach. Cook's measures of local influence are also investigated for the detection of influential cases in the data. The validity of the proportional hazards assumptions is also investigated. The residuals of Schoenfeld are examined for the possibility of being used to detect time dependence of the covariates in the proportional hazards model. Estimates to describe the nature of the time dependency computed from these residuals are presented.

The FAME yield from microalgae of two in situ transesterification methods were compared to a typical Folch et al. (1957) extraction followed by transesterification using the Hilditch et al. (1964) procedure. A method based on Park & Goins (1994), utilizing 0.5 N NaOH in methanol, then 14 % BCl3 in methanol, was found to be superior to a method based on Lepage & Roy (1986), utilizing acetyl chloride in methanol. The Park & Goins (1994) method was equivalent to the traditional method and was, therefore, selected for further study. In establishing the parameters of the method, water contents up to 0.55 mL were not found to inhibit the reaction within the maximum lipid load, conservatively assessed at ~1 mg. The reaction time and temperature required to produce a maximum FAME yield was 10 min at 90 °C for the BCl3-catalyzed reaction, while the NaOH-catalyzed reaction happened instantaneously at ambient temperature.

Nonlinear partial differential equations are difficult to solve, with many of the approximate solutions in the literature being numerical in nature. In this work, we apply the Homotopy Analysis Method to give approximate analytical solutions to nonlinear ordinary and partial differential equations. The main goal is to apply different linear operators, which can be chosen, to solve nonlinear problems. In the first three chapters, we study ordinary differential equations (ODEs) with one or two linear operators. As we progress, we apply the method to partial differential equations (PDEs) and use several linear operators. The results are all purely analytical, meaning these are approximate solutions that we can evaluate at points and take their derivatives. Another main focus is error analysis, where we test how good our approximations are. The method will always produce approximations, but we use residual errors on the domain of the problem to find a measure of error. In the last two chapters, we apply similarity transforms to PDEs to transform them into ODEs. We then use the Homotopy Analysis Method on one, but are able to find exact solutions to both equations.

Methods of efficient Monte-Carlo simulation when rare events are involved have been studied for several decades. Rare events are very important in the context of evaluating high quality computer/communication systems. Meanwhile, the efficient simulation of systems involving rare events poses great challenges. A simulation method is said to be efficient if the number of replicas required to get accurate estimates grows slowly, compared to the rate at which the probability of the rare event approaches zero. Despite the great success of the two mainstream methods, importance sampling (IS) and importance splitting, either of them can become inefficient under certain conditions, as reported in some recent studies. The purpose of this study is to look for possible enhancement of fast simulation methods. I focus on the ``level/phase process', a Markov process in which the level and the phase are two state variables. Furthermore, changes of level and phase are induced by events, which have rates that are independent of the level except at a boundary. For such a system, the event of reaching a high level occurs rarely, provided the system typically stays at lower levels. The states at those high levels constitute the rare event set. Though simple, this models a variety of applications involving rare events. In this setting, I have studied two efficient simulation methods, the rate tilting method and the adaptive splitting method, concerning their efficiencies. I have compared the efficiency of rate tilting with several previously used similar methods. The experiments are done by using queues in tandem, an often used test bench for the rare event simulation. The schema of adaptive splitting has not been described in literature. For this method, I have analyzed its efficiency to show its superiority over the (conventional) splitting method. The way that a system approaches a designated rare event set is called the system's large deviation behavior. Toward the end of gaining insight about the relation of system behavior and the efficiency of IS simulation, I quantify the large deviation behavior and its complexity. This work indicates that the system's large deviation behavior has a significant impact on the efficiency of a simulation method.

Methods of efficient Monte-Carlo simulation when rare events are involved have been studied for several decades. Rare events are very important in the context of evaluating high quality computer/communication systems. Meanwhile, the efficient simulation of systems involving rare events poses great challenges. A simulation method is said to be efficient if the number of replicas required to get accurate estimates grows slowly, compared to the rate at which the probability of the rare event approaches zero. Despite the great success of the two mainstream methods, importance sampling (IS) and importance splitting, either of them can become inefficient under certain conditions, as reported in some recent studies. The purpose of this study is to look for possible enhancement of fast simulation methods. I focus on the ``level/phase process', a Markov process in which the level and the phase are two state variables. Furthermore, changes of level and phase are induced by events, which have rates that are independent of the level except at a boundary. For such a system, the event of reaching a high level occurs rarely, provided the system typically stays at lower levels. The states at those high levels constitute the rare event set. Though simple, this models a variety of applications involving rare events. In this setting, I have studied two efficient simulation methods, the rate tilting method and the adaptive splitting method, concerning their efficiencies. I have compared the efficiency of rate tilting with several previously used similar methods. The experiments are done by using queues in tandem, an often used test bench for the rare event simulation. The schema of adaptive splitting has not been described in literature. For this method, I have analyzed its efficiency to show its superiority over the (conventional) splitting method. The way that a system approaches a designated rare event set is called the system's large deviation behavior. Toward the end of gaining insight about the relation of system behavior and the efficiency of IS simulation, I quantify the large deviation behavior and its complexity. This work indicates that the system's large deviation behavior has a significant impact on the efficiency of a simulation method.

¡@¡@The motivation of this research is to construct a Composite Index of funds which can help investors to choose funds with better ¡§future¡¨ performance. ¡@¡@The Composite Index in this thesis includes 14 kinds of indexes, such as Sharpe, Treynor, Sortino, and etc. Each kind of index is calculated by 4 different time lengths, they are 1 month, 3 months, 6 months, and 1 year. Therefore, there are 56 indexes in the Composite Index (CI). ¡@¡@Factor analysis method was used to analyze fund performance from 1998/01 to 2000/12, and the perfect weight combination to make 56 indexes become one CI was found out. In order to prove the performance of the CI in selecting funds, funds were distributed into 5 groups by their Composite Index scores every month. Therefore, we had group A to E from high CI score to low CI score. We calculated these funds¡¦ returns in the next month, and cumulated them by group from 2001/01 to 2005/12. After calculating, the cumulative returns of group A are 74.47% higher than group E, annual returns are 11.99%, and Sharpe index is 1.41. It shows that Composite Index can really distinguish the future, at least in one month, performance of stock funds. Investors are recommended to change their fund portfolios by latest CI scores once a month. Therefore, the Composite Index is more suitable for fund-of-fund manager, because their transaction costs are lower. ¡@¡@This thesis was awarded by ¡§Industry-University Cooperative Research Project¡¨ of National Science Council, and is going to be developed in PRISS system, a financial analysis software of Folion Financial Technology Co., Ltd. Therefore, a more systemized, programmed, and efficient environment for this kind of research is expectable.

Since the EPZ was set up in 1966, it has attracted foreign investment, introduced technology, opened up foreign trade and provided employment opportunities. When EPZ was first set up, it had every kind of investment-rewarding regulations, ¡§One-stop Window¡¨ administrational measures, so it could attract a lot of foreign and overseas Chinese capitals. But, after 2 years, the 68-hectare area was not enough for using. The EPZA has continuously expanded the zones up to 9 so far: the total area is about 576.81 hectares. The EPZs are export-oriented, their products are all exported, and the professions have increased step by sep, including trade, logistics and warehouses, software consulting and related industries. Its capital resources were mainly from Europe, U.S., Japan and Hong Kong, but converts into domestic mostly now, with 47.79%, the highest ratio. The Zones¡¦ industrial structure also changed from the original garments, plastics and leather with more than 55% to electronics & electric machinery with 80% the most now. The Export Processing Zone Administration (EPZA) has also coped with the era trend changes by innovating the policies and functions, such as efforts in zone transformation, deregulations, trade liberalization and industrial clustering. As the accumulations of Taiwan¡¦s capital and technology and the enormous changes both in global politics and economy, the in-zone enterprises have also changed the industrial structures and production processes, walk out the OEM shadows and walk in the ODM, so as to produce and do marketing by themselves and to expand the markets. They have transformed from labor-intensive industries of garments, plastics, leather to capital-intensive high-tech industries of IC, LCD, optics, precise tools, information software and digital content and so on. When Taiwan¡¦s producing costs are rising little by little, the enterprises form all over the world are moving to the mainland China and the booming southeastern Asian countries to reduce costs and grip the markets. It is the question how to mold the more suitable investment environment to lure the Taiwanese enterprises come back home and the foreign investment turn around to Taiwan. This study made the in-depth interviews and did the questionnaires to scholars, government agencies and the industrial circle for over 30 copies. This study also analyzes the key factors of attracting investment by AHP hierarchical analysis method. The results revealed that, firstly, the enterprises¡¦ interactions and the neighboring of customers, the benefits of industrial cluster and the completion of supply chain of down and up stream, are vital to attract investment. Secondly, the completion of the in-zone infrastructure is the most important key factor for attracting investment. Thirdly, The EPZA ¡¥s favorable measures and administrative efficiency are also the prior considering factors. This study suggested that the administration units should upgrade the functions of their in-zone administrative steps, build the modern facilities of the infrastructure and attach importance to the living functions, strengthen the cooperation between industries, governments and schools, and propagate the investment-rewarding measures offered by governments to the enterprises, so that the governments¡¦ good policies can truly do favors to the in-zone enterprises, and thus construct the zones¡¦ better investment environment and attract investment.

Entrepreneurship is a booming phenomenon in Taiwan and there are quite many research focusing on this domain continually. This research¡¦s aim is that entrepreneurship not only can create economical profits, but also is a link between social networks and interpersonal relationships. Emergence of entrepreneurial opportunities, change of social environment, establishment and maintenance of interpersonal and organization¡¦s network, knowledge learning and accumulation, are some of the indispensable factors for a successful entrepreneurship. In addition to the above, Taiwan depends heavily on its foreign trade; trade related industry holds quite a big contribution to the Taiwan economic growth. Therefore, this research adopts a trading company in gardening industry for case study, so as to know the problems and solutions, also to find out if there are any hidden implications from the story. This research is written in a narrative analysis method just to reorganize and annotate the hidden context throughout the presentation of episodes in one¡¦s life story. It is a story of an entrepreneur that grows out of nothing. Key words: Entrepreneurship, emergence of opportunities, interpersonal network, learning, gardening, narrative analysis method.

The statistical results of the field of Chinese FBG sensing technology were visualized by using patent analysis method and parsed the technology background of FBG sensors. Several statistics in 17 parameters system were completed, which include the current status, R&D teams, and application fields of the patents of FBG sensors. The statistical results indicate that the FGB sensing technology are currently in the development of the technology life cycle; the R&D focus concentrates on the sensor designing and sensor array system; college labs account the most R&D teams; the application field includes the composite and concrete structure monitoring, the performance monitoring of smart material, power industry, pharmaceutical industry, and chemical industry. With the gradual increase in R&D, gradually mature in technology, the FBG will in the mature of the technology life cycle; and wavelength demodulation device will be an significant direction of development of FBG sensing technology.

No description available.