Essays in the pure theory of international trade : the effects of economies of scale and technological innovation on the pattern of trade

Dollar, David Edward Elgar 1984 국내박사

Learning from local image regions

Dollar, Piotr University of California, San Diego 2007 해외박사(DDOD)

A trend in computer vision over the last decade or so has been to describe the statistics and content of images in terms of local image regions, i.e., image patches. Applications have included object detection, scene recognition, texture classification and image categorization. Local patch based representations have the advantage that they are robust to global transformations, occlusion, clutter, object and image variation, and so on, while retaining rich information about image content. This is the case even when global information relating the relative position of patches is not used, as in so called "bags of words" approaches. Furthermore, in the supervised learning framework where labeled images are a source of data, characterizing images using patches means a single image can provide a large number of patches for training. These properties suggest local patch based representations should continue to find expanded use in computer vision. In this dissertation we show the application of patch based methods to three domains for which traditionally more global approaches have been used. First we show how the classic problem of edge detection can be posed as a series of patch by patch decisions that can be solved in a supervised learning framework. We show the application of this approach to a number of specific domains such as mouse boundary detection and road detection. Second, we show how modeling object warps and highly non-linear image transformations can again be done locally, thus avoiding computational challenges and the scarcity of data typically associated with these problems. For example, our approach is able to learn eye motion and out-of-plane rotation of a teacup from sparse data. Third, we extend the notion of local regions from 2D to 3D, i.e. from patches to cuboids, in order to model the content of video. We show applications to behavior recognition in a number of domains including human activity and mouse behavior. The methods we introduce here advance the state of the art and have the potential to be useful in a broad range of applications in computer vision. Our approach to edge detection currently outperforms all competing approaches for gray scale edge detection and comes in close second for color edge detection on the well established Berkeley Segmentation Dataset. We hope it will play a similar role as Canny edge detection but for highly textured, real world images. Our approach to modeling object warps locally showed dramatic improvements over previous such methods, and helped solidify the theoretical foundation of nonlinear manifold learning. Finally, our cuboids formalism is simple yet powerful, and has already been utilized in two vision systems. It has the potential to serve as the basis for a broad range of methods for describing the contents of video. Overall, our contribution has been to help establish the importance of patch based approaches and to expand our understanding of a fundamental aspect of computer vision.

The class of 1990: A longitudinal study of a freshman cohort at Texas A&M University-Kingsville

Dollar, Susan Texas A&M University 2005 해외박사(DDOD)

Extensive research has been conducted on college student retention and graduation and many studies have found certain characteristics to be predictive of successful completion of college. However, few studies have focused on a target population which is primarily Hispanic. This study examined the 1990 entering freshmen class of students at Texas A&M University-Kingsville (TAMUK), of which more than 68% were Hispanic. An attempt was made to examine characteristics that would predict success, defined as graduation from TAMUK. Data derived from institutional records and the Texas Higher Education Coordinating Board were examined using descriptive statistics and stepwise multiple logistic regression. Pre-college characteristics studied included age, gender, ethnicity, marital status, high school GPA (Grade Point Average), high school class rank, high school of origin, county of origin, and American College Test (ACT) and the Scholastic Assessment Test (SAT) scores. In-college characteristics studied included residency status, admission status, enrollment status, number of hours enrolled fall 1990, college major, the Texas Academic Skills Program (TASP) scores, developmental courses, semester GPA's, academic standing, and finally, attrition, transfer or graduation status. The fall 1990 entering students were evenly divided between males (53.4%) and females (46.6%), were young (79%) were age 19 or less), single (91.4%), and Hispanic (68.2%). Almost half (46%) of the students came from high schools within 50 miles of Kingsville. The mean college entrance exam scores (ACT = 16.76 and SAT = 766) were well below the national means of 21 and 999, respectively. Of the 1106 entering freshmen, 307 (27.7%) graduated from TAMUK within the 10 years under study. An additional 490 (44.3%) transferred to other state institutions, and 309 (27.9%) dropped out of TAMUK and did not enroll in any other state college or university. The fall-to-spring attrition rate was only 16.5%; however, the fall-to-fall attrition rate was 50.0% at the end of the first year. Stepwise multiple logistic regression (backward) analysis revealed that only high school GPA and the ACT composite score were statistically significant predictors of graduation.

A predictive study of selected biological, social, behavioral and environmental risk factors for low birth weight infant births to African-American women in Mississippi

Dollar, Carolyn Blalock The University of Mississippi Medical Center 2005 해외박사(DDOD)

There is considerable racial disparity between African-American and White low birth weight (LBW) infant births in Mississippi. Low birth weight statistics for Mississippi per 1000 live births for the year 2000 were: total 10.3, White infants 7.4, and African-American infants 13.8. National LBW statistics were: total 7.6, White infants 5.7, and African-American infants 13.1, and follow a similar pattern in terms of racial disparity. Birth weight is the greatest predictor of infant health. The purpose of this study was to identify and describe selected biological factors (age, parity, prior fetal loss, preexisting medical risk factors, and complications of pregnancy), social factors of marriage (proxy for social support), and education (proxy for socioeconomic status), behavioral factors (alcohol use, smoking, prenatal care, and weight gain during pregnancy), and environmental factors (unemployment, mean income, population, racial mix, and number of health care providers) for each Mississippi county, either alone or in combination, that were most predictive of LBW infant births in African-American women in Mississippi. The study sample of 3,950 live births was obtained from certificates of live births for all African-American singleton births in the year 2000 from the Mississippi State Department of Health (MSDH) Bureau of Vital Statistics. Environmental factors were obtained from the 2000 U.S. Census Records and the MSDH for Mississippi's eighty-two counties. A predictive correlational design with descriptive statistics, Pearson correlation coefficients, and forward stepwise logistic regression were used as methods of analysis. A model of risk factors most predictive of LBW infant births included: weight gain <25 pounds; lack of prenatal care below 50% expected visits and greater than 110% expected visits; pregnancy complications (any level); prior fetal loss (one or two); and smoking during pregnancy. Environmental factors significantly associated with LBW infant births were: unemployment, racial mix, and number of primary care providers in each county. A predictive model of risk factors for LBW infant births in African-American women in Mississippi will increase understanding and assist with development of culturally sensitive programs of assessment, intervention, education, prevention, and evaluation in order to improve infant health.

High intensity, high contrast laser solid interactions with short pulses

Dollar, Franklin Jon University of Michigan 2012 해외박사(DDOD)

This thesis describes experimental discoveries related to laser-based ion acceleration from thin foils and the production of high brightness x-rays from high order harmonic generation. High power femtosecond lasers are ideally suited for use as tabletop particle accelerators since their short pulse duration enables very high intensities to be generated at high repetition rates from a compact laser. However, if laser pulse energy arrives before the main short pulse, it can interact with the target to cause ablation making high intensity investigations of laser-solid interactions difficult. In the following experiments, the laser pulse-to-pedestal contrast was improved by 15 orders of magnitude out to nanosecond timescales, allowing for excellent control over the interaction of a short pulse with solid density material. A sharply-rising laser pulse with 50 TW of power was focused to a 1.2 micron focal spot, achieving intensities over 10<super>21</super>Wcm<super>– 2</super>. Protons accelerated due to sheath acceleration were studied in ultrathin targets. By sculpting the plasma density using shaped ultrafast pulses, control over the proton and ion spectra was also demonstrated. Finite spot effects from circular polarized laser pulses produced efficient acceleration for ultrathin foils, which resulted from the efficient conversion of laser light into high energy electrons. Finally, as the laser pulse drives the critical electron density relativistically, harmonics of the driving laser are produced. Harmonics up to order 60<super>th</super> were observed. It was observed that for a plasma scale length beyond a threshold value, parametric instabilities strongly modulated the harmonic spectra. Numerical simulations were performed to support the physical interpretation.

Migration, Worker Empowerment, and the Development of State Agricultural Labor Regimes: An Historical Comparative Case Study of California and North Carolina (1880-2022)

Dollar, Nathan Tilghman ProQuest Dissertations & Theses The University of 2022 해외박사(DDOD)

In this dissertation, I compare the historical development of state agricultural labor policies and practices in California, a traditional migrant destination, and North Carolina, a new migrant destination. I identify the factors which shaped the emergence of a more protective environment in California and a more precarious environment in North Carolina. I then identify how these contrasting state contexts affect the health and well-being of Latino migrant farmworkers in each state today. Drawing on archival data, field observations, and 37 interviews with farmworkers in California (n=22) and North Carolina (n=15), I develop a labor regime model to explain how the different state sociopolitical cultures developed, and how they impinge on the health and well-being of migrant farmworkers and their families. I demonstrate how the development of agricultural labor regimes is intimately tied to systems of agricultural production, migration patterns, and their implications for worker empowerment. Foreign-born and domestic migrants have been recruited to work on California’s industrial farms since the late nineteenth century. These migrant farmworkers have engaged in cross-ethnic collective resistance against employer abuse for 130 years. Their efforts have been instrumental in constructing the more protective agricultural labor regime we observe in California today. In North Carolina, the harsh conditions associated with the disjointed system of tenant farming, sharecropping, and plantation production in North Carolina served as a deterrent for potential migrants and pushed many Black agricultural workers out. This isolation from outsiders, combined with the mass exodus of Black agricultural workers – whose solidarity and strong labor consciousness made them the most likely to organize and collectively resist – constrained the potential for worker empowerment until the 1990s. This resulted in the consolidation of employers’ hegemony over the North Carolina’s political apparatus and the precarious labor regime we observe in the state today. I demonstrate how these state agricultural labor regimes have implications for the health and well-being of migrant farmworkers and their families. These findings enhance our understanding of how migrants’ labor market incorporation is embedded in the sociopolitical histories of the places where migrants live and work.