In this thesis a machine vision based indoor navigation system is presented. This is achieved by using rotationally independent optimized color reference labels and a geometrical camera calibration model which determines a set of camera parameters. All reference labels carry one byte of information (0 to 255), which can be designed for different values. An algorithm in Matlab has been developed so that a machine vision system for N number of symbols can recognize the symbols at different orientations. A camera calibration model describes the mapping between the 3-D world coordinates and the 2-D image coordinates. The reconstruction system uses the direct linear transform (DLT) method with a set of control reference labels in relation to the camera calibration. The least-squares adjustment method has been developed to calculate the parameters of the machine vision system. In these experiments it has been demonstrated that the pose of the camera can be calculated, with a relatively high precision, by using the least-squares estimation.

This thesis proposes an adaptive control approach to regulate the hydrogen feed of a fuel cell. The goal of the controller is to maintain the so-called hydrogen excess ratio, defined as the ratio between the hydrogen fed to the cell stake and those consumed in the stake, at a desired level when the fuel cell is under load variation. Maintaining the hydrogen excess ratio at an appropriate level would avoid hydrogen starvation, which is crucial for slowing degeneration of the fuel cell membranes and prolonging the life of the cell stake. The control approach we propose is based on the receding horizon linear quadratic optimal control algorithm with an on-line turning scheme which updates the plant model according to real-time measurement. To ease the computational complexity and make real-time turning realizable, we adopt a simple autoregressive with external disturbance (ARX) model to approximate the complicate chemical/electrical process of the fuel cell. The proposed adaptive control approach is implemented on an experimental platform. The experimental results show that the proposed control works with reasonably good performance.

This paper studies the role of fluctuation of the aggregate consumption-wealth ratio(cay) for predicting Taiwan stock return. The effect of cay on U.S. stock return has been recently confirmed by Lettau and Ludvigson (2001) with a two stage method. In the first step, estimate the ratio used a dynamic least square(DLS) technique. Second, to investigate the performance of cay, they use in-sample and out-of-sample test. In this paper, we follow the method which Lettau and Ludvigson(2001) use to examine the predictability of cay. Using quarterly market data from 1998 to 2010, we find cay is strong predictors of excess return in out-of sample test. We also find that this ratio is a better forecaster of future returns at intermediate horizons compared to short time.

This paper is concerned with the problem of finding an experimental design for discrimination between two rival models and for model robustness that minimizing the maximum bias simultaneously in binary response experiments. The criterion for model discrimination is based on the $T$-optimality criterion proposed in Atkinson and Fedorov (1975), which maximizes the sum of squares of deviations between the two rival models while the criterion for model robustness is based on minimizing the maximum probability bias of the two rival models. In this paper we obtain the optimum designs satisfy the above two criteria for some commonly used rival models in binary response experiments such as the probit and logit models etc.

Abstract To improve the application of Point Interpolation Method (PIM) in Element Free Galerkin Method (EFG) is the aim of this study. The trait of EFG is using overlap of influence domain between different nodes to construct discretization nodes¡¦ connection. EFG just uses nodal data, but not element. For constructing shape function, EFG has two types of methods, Fitting and Interpolation. Fitting uses Moving Least Square Method (MLS). MLS-EFG has stable effect on numerical analysis; however, users who use it need to choose more numerical parameters and do more computation. Besides, users can not apply boundary conditions directly when using MLS-EFG. Interpolation method applies nodal coordinates to proceed computation, and it called PIM. Boundary conditions could be used directly and less computation is needs while using PIM. However, the coefficient of interpolation function of sample is singular. This study tries to construct Coordination Point Interpolation Method. It owns advantages of both methods that mentioned above, and extra numerical parameters are not needed. It applies the notion of influence domain of MLS-EFG, then search correlative efficient nodes which are contained in near field of sample. The correlative efficient nodes make up matrix that con cause inverse matrix. In addition, via numerical simulations, it shows that CPIM has excellent convergence and accurate solution, and is better that MLS-EFG.

Business cycle is an important issue for economist. Because the fluctuations of product and employment have deep influences on people¡¦s life and social stability, almost every government tries to reduce the volatility of national business cycles. If we want to make it, we must realize it first. Since countries communicate with each other more frequently in recent decades, the volatility of national product cycles is not only influenced by domestic economic variables but also foreign ones. The purpose of this paper is to analyze the effect of transmission mechanism on international business cycles synchronization (BCCs). The major purpose of this paper is to analyze the effect of international vertical specialization on BCCs among industrial countries. There are two kinds of effect. One is indirect effect. Vertical specialization happens between industrial countries and developing countries, so it can reduce bilateral trade intensity among industrial countries. Through this way, BCCs among industrial countries will reduce. The other one is direct effect. Vertical specialization changes the economic structure of industrial countries. Industrial countries can focus on product development and market research. This kind of economic structure is less capital intensive, which lessens the effect of common shocks to industrial country¡¦s business cycles. BCCs among industrial countries will reduce. We measured the effect using the data from G6 (Canada, France, Germany, Japan, UK and US). The result is consistent with our inferance. Vertical specialization can reduce BCCs by reducing bilateral trade intensive among industrial countries. There is a negative relation between Vertical specialization and BCCs among industrial countries.

The aim of this thesis is to identify factors that affect the university students&rsquo / acceptance of multi-purpose Smart Cards. The findings of this study will be beneficial to facilitate the use of Smart-Card enabled system both n universities and in other institutions which either have these systems in use or plan to invest on these systems in the future. The research methodology employed within this study is based on quantitative methods. A survey instrument comprising 51 5-point Likert-type questions has been developed and applied to 207 university Middle East Technical University students. The data collected has been analyzed using Exploratory Factor Analysis to categorize factors having items. According to analysis results, the data classified under 5 factors / Perceived Usefulness, Perceived Ease of Use, Behavioral Intention, Anxiety, and Technological Complexity. Then, the relations between these 5 factors identified and a measurement model was created. For assessing the proposed model Discriminant and Convergent Validity scores were calculated by Confirmatory Factor Analysis. Then, Structural Equation Modeling was conducted with Partial Least Squares for validating the model&rsquo / s estimated influence. The study has shown that the main Technology Acceptance Model constructs fit for determining the university students&rsquo / intention of Smart Card usage except for Perceived Ease of Use over Behavioral Intention. Moreover, study showed that Anxiety and Technological Complexity were the external factors that have effect on willingness of using multi-purpose Smart Cards. If students have Anxiety, this affects their perception of easiness of the system and it has negative indirect effect on the perceived usefulness and direct effect on intention. Technological Complexity is another factor which has direct affect on the perception of easiness and usefulness and intention.

In this work, it is of interest to study the effects of thrombocytosis and microvascular anastomosis to that of vascular thrombosis. Here the splenectomy is used to the experimental subjects to have caused thrombocytosis first and later have the microvascular anastomsis on the divided vessel. Observations on different variables which may be related to vascular thrombosis are obtained during the three stages of the experiments, there are pre-splenectomy, post-splenectomy but before vascular division, and after the vascular anastomsis. These variables include platelet number, RBC count, Nitric oxide(nitrite and nitrate), femoral artery perfusion unit, hematocrit, reticulated-platelet number, and activated-platelets(p-selectin, CD62p). The experimental units are divided into thrombocysis group and control group. Basic statistical analysis are performed first to see the differences between the two groups for the above mentioned variables, and later try to use the profile analysis in multivariate analysis to understand the changes of these variables in different stages as well as their possible relationships with the cause of vascular thrombosis. Finally, classical regression and robust regression analysis for finding outliers are also used to see whether certain characteristics of those outliers have something to do with vascular thrombosis.

To significantly improve the early detection of cervical precancers and cancers, LightTouch™ is under development by SpectRx Inc.. LightTouch™ identifies cancers and precancers quickly by using a spectrometer to analyze light reflected from the cervix. Data from the spectrometer is then used to create an image of the cervix that highlights the location and severity of disease. Our research is conducted to find the appropriate models that can be used to generate map-like image showing disease tissue from normal and further diagnose the cervical cancerous conditions. Through large work of explanatory variable search and reduction, logistic regression and Partial Least Square Regression successfully applied to our modeling process. These models were validated by 60/40 cross validation and 10 folder cross validation. Further examination of model performance, such as AUC, sensitivity and specificity, threshold had been conducted.

In phylogenetic analysis, an important issue is to construct the confidence region for gene trees from DNA sequences. Usually estimation of the trees is the initial step. Maximum likelihood methods are widely applied but few tests are based on distance methods. In this thesis, we propose a new test based on balanced minimum evolution. We first examine the normality assumption of pairwise distance estimates under various model misspeci cations and also examine their variances, MSEs and squared biases. Then we compare the BME method with the WLS method in true tree reconstruction under different variance structures and model pairs. Finally, we develop a new test for finding a confidence region for the tree based on the BME method and demonstrate its effectiveness through simulation.