This thesis studies the problem of frequentist model averaging over a set of multiple $\epsilon$-support vector regression (SVR) models, where the support vector machine (SVM) algorithm was extended to function estimation involving continuous targets, instead of categorical ones. By assigning weights to a set of candidate models instead of selecting the least misspecified one, model averaging presents a strong alternative to model selection for tackling model uncertainty. Not only do we describe the construction of smoothed BIC/AIC model averaging weights, but we also propose a Mallows model averaging procedure which selects model weights by minimizing Mallows' criterion. We conduct two studies where the set of candidate models can either include or not include the true model by making use of simulated random samples obtained from different data-generating processes of analytic form. In terms of mean squared error, we demonstrate that our proposed method outperforms other model averaging and model selection methods that were tested, and the gain is more substantial for smaller sample sizes with larger signal-to-noise ratios. / Thesis / Master of Science (MSc)

This dissertation covers three levels of issues and solutions dealing with unbalanced and/or nonlinear situations in power electronic systems, namely power converter level, power converter system level, and large-scale power electronics system level. At power converter level, after review of traditional PWM methods, especially two-dimensional space vector modulation schemes, three-dimensional space vector modulation schemes are proposed for four-legged voltage source converters, including inverters and rectifiers. The four-legged power converters with three-dimensional space vector modulation schemes have a better DC link voltage utilization and result in a low distortion. It is an effective solution to provide the neutral point for a three-phase four-wire system and to handle the neutral current due to unbalanced load or source and nonlinear loads. Comprehensive design guidelines for a four-legged inverter are presented. The four-legged rectifier is also presented which allows not only fault tolerant operation, but also provides the flexibility of equal resistance, equal current, or equal power operation under unbalanced source. Average large-signal models of four-legged power converters in both the a-b-c and d-q-o coordinates are derived. Small signal models are obtained in the d-q-o rotating coordinates. Voltage control loops are designed in the d-q-o coordinates for a high power utility power supply. Performance is studied under various load conditions. At the power converter system level, the load conditioner concept is proposed for high power applications. A power converter system structure is proposed which consists of a high-power low-switching frequency main inverter and a low-power high-switching frequency load conditioner. The load conditioner performs multiple functions, such as active filtering, active damping, and active decoupling with a high current control bandwidth. This hybrid approach allows the overall system to achieve high performance with high power and highly nonlinear loads. At the large-scale power electronics system level, the nonlinear loading effect of load converters is analyzed for a DC distribution system. Two solutions to the nonlinear loading effect are presented. One is to confine the nonlinear load effect with the sub-converter system, the other is to use a DC bus conditioner. The DC bus conditioner is the extension of the load conditioner concept. / Ph. D.

This dissertation addresses three important open questions in the context of generating random vectors having discrete support. The first question relates to the "NORmal To Anything" (NORTA) procedure, which is easily the most widely used amongst methods for general random vector generation. While NORTA enjoys such popularity, there remain issues surrounding its efficient and correct implementation particularly when generating random vectors having denumerable support. These complications stem primarily from having to safely compute (on a digital computer) certain infinite summations that are inherent to the NORTA procedure. This dissertation addresses the summation issue within NORTA through the construction of easily computable truncation rules that can be applied for a range of discrete random vector generation contexts. The second question tackled in this dissertation relates to developing a customized algorithm for generating multivariate Poisson random vectors. The algorithm developed (TREx) is uniformly fast—about hundred to thousand times faster than NORTA—and presents opportunities for straightforward extensions to the case of negative binomial marginal distributions. The third and arguably most important question addressed in the dissertation is that of exact nonparametric random vector generation on finite spaces. Specifically, it is wellknown that NORTA does not guarantee exact generation in dimensions higher than two. This represents an important gap in the random vector generation literature, especially in view of contexts that stipulate strict adherence to the dependency structure of the requested random vectors. This dissertation fully addresses this gap through the development of Maximum Entropy methods. The methods are exact, very efficient, and work on any finite discrete space with stipulated nonparametric marginal distributions. All code developed as part of the dissertation was written in MATLAB, and is publicly accessible through the Web site https://filebox.vt.edu/users/pasupath/pasupath.htm. / Ph. D.

International malaria control initiatives such as the Roll Back Malaria Initiative (RBM) and the Medicines for Malaria Venture (MMV) mobilize resources and spur research aimed at vector control as well as the treatment and eventual eradication of the disease. These efforts have managed to reduce incidence of malaria by an estimated 37% worldwide since 2000. However, despite the promising success of control efforts such as these, the World Health Organization reports a staggering 438,000 deaths from malaria in 2015. The continuing high death toll of malaria as well as emerging insecticide and antimalarial drug resistance suggests that while encouraging, success in reducing malaria incidence may be tenuous. Current vector control strategies are often complicated by ecological and behavioral heterogeneity of vector mosquito populations. As an additional obstruction, mosquito genomes are highly plastic as evidenced by the wealth or chromosomal inversions that have occurred in this genus. Chromosomal inversions have been correlated with differences in adaptation to aridity, insecticide resistance, and differences in resting behavior. However, a good understanding of the molecular mechanisms for inversion generation is still lacking. One possible contributor to inversion formation in Anopheles mosquitoes includes repetitive DNA such as transposable elements (TEs), tandem repeats (TRs) and inverted repeats (IRs). This dissertation provides physical maps for two important malaria vectors, An. stephensi and An. albimanus (Ch.2 and Ch. 3) and then applies those maps to the identification of inversion breakpoints in malaria mosquitoes. Repeat content of each chromosomal arm and the molecular characterization of lineage specific breakpoints is also investigated (Ch. 2 and Ch.4). Our study reveals differences in patterns of chromosomal evolution of Anopheles mosquitoes vs. Drosophila. First, mosquito chromosomes tend to shuffle as intact elements via whole arm translocations and do not under fissions or fusions as seen in fruitflies. Second, the mosquito sex chromosome is changing at a much higher rate relative to the autosomes in malaria mosquitoes than in fruit flies. Third, our molecular characterization of inversion breakpoints indicates that TEs and TRs may participate in inversion genesis in an arm specific manner. / Ph. D. / Malaria is a complex and devastating disease vectored by the bite of a female Anopheles mosquito. This disease claimed an estimated 438,000 lives in 2015. The mobilization of funding and resources as part of global malaria eradication initiatives have reduced the global incidence of malaria by 37% in the last 15 years. Deaths from malaria are also 60% lower vs. the year 2000. These promising gains are threatened by the ability of Anopheles mosquitoes to adapt in the face of malaria control efforts. Anopheles mosquito chromosomes are known to be highly plastic, as evidenced by numerous chromosomal inversions. Recent years have seen increases in insecticide resistance, and behavioral change in mosquito populations that allow them to avoid insecticides and remain prolific vectors of disease. This ability of mosquito vectors to adapt threatens to unravel recent progress towards a malaria free world. The projects presented in this dissertation explore mechanisms of chromosomal evolution, specifically the potential role of repetitive DNA in the generation of chromosomal inversions. The exploration of chromosomal inversions was facilitated by the creation of physical maps for Anopheles species. Prominent malaria vectors An. stephensi andAn. albimanus were physically mapped in Chapter 2 and Chapter 3 respectively. In chapter 1 and chapter 3 physical maps are utilized for the identification of chromosomal inversion breakpoints using 2 species (Ch. 2) and many species (Ch. 4). Repeat content was quantified along each chromosomal arm (Ch 2,4) and in inversion breakpoint regions (Ch 3). This dissertation presents physical maps for two important malaria species that have been applied to the study of chromosomal evolution and will also serve as community tools for further study of malaria mosquitoes. Our work on chromosomal evolution has revealed the Anopheles chromosomes tend to undergo translocations as intact elements and do not under fissions and fusions as seen in fruitflies. We also find that the malaria mosquito sex chromosome changes much more rapidly relative to the autosomes than in fruitflies. Additionally, repetitive DNA including transposable elements (TEs) and tandem repeats (TRs) may be encouraging chromosomal inversions but with differing roles on different chromosomal arms.

This study seeks to investigate weather-related flight delay trends at 13 United States airports. Flight delay data were collected from 2004-2019 and normalized by airport operations data. Using Support Vector Regression (SVR), visual trends were identified. Further analysis was conducted by comparing all four meteorological seasons through computing 95% bootstrap confidence intervals on their means. Finally, precipitation and snowfall data were correlated with normalized delays to investigate how they are related. This study found that the season with the highest normalized delay values is heavily dependent upon location. Most airports saw a decrease in the SVR line at some point since 2004, but have since leveled off. It was also discovered that while precipitation trends are not changing drastically, delay variability has decreased at many airports in the last 10 years, which may be indicative of more effective mitigation strategies.

Every year three million deaths are attributed to malaria, of which one-third are of children. Malaria is a vector-borne disease, where a mosquito acts as the vector that transmits the disease. In the last few years, computer simulation based models have been used effectively to study the vector population dynamics and control strategies of vector-borne diseases. Typically, these models use ordinary differential equations to simulate the spread of malaria. Although these models provide a powerful mechanism to study the spread of malaria, they have several shortcomings. The research in this thesis focuses on creating a simulation model based on the framework of cellular automata, which addresses many shortcomings of previous models. Cellular automata are dynamical systems, which are discrete in time and space. The implementation of the model proposed can easily be integrated with EpiSims/TRANSIMS. EpiSims is an epidemiological modeling tool for studying the spread of infectious diseases; it uses social contact network from TRANSIMS (A Transport Analysis and Simulation System). Simulation results from the prototype implementation showed qualitatively correct results for vector densities, diffusion and epidemiological curves. / Master of Science

Inflation targeting is a monetary policy regime, characterized by public announcement of official target ranges or quantitative targets for price level increases and by explicit acknowledgement that low inflation is the most crucial long-run objective of the monetary authorities. There are three prerequisites for inflation targeting: 1)central bank independence,2)having a sole target,3)existence of stable and predictable relationship between monetary policy instruments and inflation.In many developing countries, the use of seigniorage revenues as an important source of financing public debts, the lack of commitment to low inflation as a primary goal by monetary authorities, considerable exchange rate flexibility, lack of substantial operational independence of the central bank or of powerful models to make domestic inflation forecasts hinder the satisfaction of these requirements. This study investigates the applicability of inflation targeting to the Turkish economy. Central bank independence in Turkey has been mainly hindered by "fiscal dominance" through monetization of high budget deficits. In addition, although serious steps have been taken recently under a new law to have an independent central bank, such as formal commitment to the achievement of price stability as the primary objective and the prohibition of credit extension to the government, the central bank does not satisfy independence criteria due to the problems associated with the appointment of the government and the share of the Treasury within the bank. Having a sole inflation target was hindered by the existence of fixed exchange rate system throughout the years. However, in February 2001, Turkey switched to a floating exchange rate regime, which is important for a successful inflation-targeting regime. Having a sole target within the system has also been supported by the new central bank law, which gives priority to price stability and supports any other objective as long as it is consistent with price stability. In this thesis, an empirical investigation has been made in order to assess the statistical readiness of Turkey to satisfy the requirements of inflation-targeting by making use of vector autoregressive (VAR) models. The results suggest that inflation is an inertial phenomenon in Turkey and money, interest rates and nominal exchange rates innovations are not economically and statistically important determinants of prices. Most of the variances in prices are explained by prices themselves. According to the VAR evidence, the direct linkages between monetary policy instruments and inflation do not seem to be strong, stable, and predictable. As a result, while the second requirement of the inflation-targeting regime seems to have been satisfied, there are still problems associated with the central bank independence and the existence of stable and predictable relationship between monetary policy instruments and inflation in Turkey. / Master of Arts

No / An algorithm called VPAStab is given for the acceleration of convergence of a sequence of vectors. It combines a method of vector-Padé approximation with a successful technique for stabilisation. More generally, this algorithm is designed to find the fixed point of the generating function of the given sequence of vectors, analogously to the way in which ordinary Padé approximants can accelerate the convergence of a given scalar sequence. VPAStab is justified in the context of its application to the solution of a large sparse system of linear equations. The possible breakdowns of the algorithm are listed. Numerical experiments indicate that these breakdowns can be classified either as pivot-type (type L) or as ghost-type (type D).

This research simulates pre-retirement investment scenarios for agricultural producers. Thirty-two investment scenarios are examined, with each scenario differing with respect to retirement vehicle, investment strategy of the producer, and the use of a cash margin for reinvestment in the operation versus prepaying term debt (cash preference). The retirement vehicles included in this study are Individual Retirement Accounts (IRAs), Simplified Employee Pension Plans (SEPs), and 401(k) plans. Investment strategies reflect the producer's preference for investing in conservative, balanced, or aggressive assets, or a combination of these assets. Further, these scenarios are examined for three methods of capitalization: Case I- an operation with a 50 percent debt/asset ratio; Case II - an operation with a 65 percent debt/asset ratio; Case III - an operation with a 65 percent debt/asset ratio with a majority of the farm land being leased. The analytical model simulates the annual cash flows of a commercial agricultural operation for each investment scenario over a 30-year period. Stochastic rates of return, generated using a vector-autoregressive (VAR) model, are incorporated into the simulation model. Each scenario is replicated 100 times using different vectors of stochastic rates of return. Results show investment in retirement vehicles does not significant reduce ending farm assets, regardless of investment strategy or cash preference of the producer. Use of retirement vehicles does have a significant positive impact on ending net worth for the producer. IRAs are not significant investment tools for producers (or spouses) who are participants in another qualified retirement plan. Investment strategy has a major impact on ending net worth. Aggressive and dynamic (aggressive to conservative as retirement approaches) investment strategies dominate conservative and balanced strategies. Use of cash margin to prepay debt has no advantage over reinvesting in the farm. Retirement vehicles greatly improve the probability of meeting estimated family living needs during retirement, and generate greater diversity and liquidity of the retirement portfolio. Further, retirement vehicles are more important for producer with highly-leveraged operations and for producers who lease a majority of their assets. / Ph. D.

Visualization of trivariate functions and vector fields with three components in scientific computation is still a hard problem in compute graphic area. People build their own visualization packages for their special purposes. And there exist some general-purpose packages (MatLab, Vis5D), but they all require extensive user experience on setting all the parameters in order to generate images. We present a simple package to produce simplified but productive images of 3-D vector fields. We used this method to render the magnetic field and current as solutions of the Ginzburg-Landau equations on a 3-D domain.