abstract: This article proposes a new information-based subdata selection (IBOSS) algorithm, Squared Scaled Distance Algorithm (SSDA). It is based on the invariance of the determinant of the information matrix under orthogonal transformations, especially rotations. Extensive simulation results show that the new IBOSS algorithm retains nice asymptotic properties of IBOSS and gives a larger determinant of the subdata information matrix. It has the same order of time complexity as the D-optimal IBOSS algorithm. However, it exploits the advantages of vectorized calculation avoiding for loops and is approximately 6 times as fast as the D-optimal IBOSS algorithm in R. The robustness of SSDA is studied from three aspects: nonorthogonality, including interaction terms and variable misspecification. A new accurate variable selection algorithm is proposed to help the implementation of IBOSS algorithms when a large number of variables are present with sparse important variables among them. Aggregating random subsample results, this variable selection algorithm is much more accurate than the LASSO method using full data. Since the time complexity is associated with the number of variables only, it is also very computationally efficient if the number of variables is fixed as n increases and not massively large. More importantly, using subsamples it solves the problem that full data cannot be stored in the memory when a data set is too large. / Dissertation/Thesis / Masters Thesis Statistics 2017

Chez les organismes pluricellulaires, la diversité morphologique observée à tous les niveaux est frappante, que ce soit au niveau de la taille, des couleurs, ou de la forme des individus, et des différentes parties qui les composent. La sélection naturelle, ainsi que les contraintes développementales, influencent l’évolution de ces différents attributs, sur le court terme comme sur le long terme. Le modèle choisi ici pour étudier l’évolution des formes est le grain de pollen des plantes à fleurs, qui présente une très grande diversité morphologique. Nous nous sommes focalisés sur un caractère morphologique, les apertures, qui sont des structures de la paroi du grain de pollen impliquées dans la survie et la reproduction. Nous avons étudié l’évolution des apertures à grande échelle taxonomique chez les angiospermes, et nous montrons que s’il existe de nombreuses variations, deux types principaux dominent : un pollen à une aperture chez les Monocotylédones et les angiospermes divergeant à la base, et un pollen à trois apertures chez les Eudicotylédones. En étudiant la dominance du pollen à trois apertures, nous avons pu montrer que la stase chez les Eudicotylédones était vraisemblablement due à une sélection stabilisante plutôt qu’à des contraintes développementales. Nous avons également montré, grâce à l’utilisation de mutants de la plante modèle Arabidopsis thaliana, qu’un nombre d’apertures élevé est défavorable face à un stress osmotique, ce qui pourrait suggérer que les pollens triaperturés représentent un bon compromis entre survie et germination. Enfin, nous nous sommes intéressés à un type particulier de pollen possédant de nombreuses apertures, dont nous avons étudié la distribution chez les angiospermes. L’apparition de ce type de pollen est récurrente, mais il n’est que rarement fixé à grande échelle taxonomique, suggérant l’existence d’un mécanisme de sélection interphylétique éliminant ce type de pollen sur le long terme. / Multicellular organisms are morphologically very diverse at every scale, regarding size, color, and shape of individuals and of their different parts. Natural selection and developmental constraints influence evolution of these characteristics, on the short term as well as on the long term. The model chosen here to study form evolution is the pollen grain of flowering plants, which is very diversified morphologically. We focused on specific features called apertures, which are structures of the pollen wall involved in survival and reproduction. We studied aperture evolution at large taxonomic scale in angiosperms, and we showed that in spite of variations, two main pollen types dominate: a pollen with one aperture in Monocots and early diverging angiosperms, and a pollen with three apertures in Eudicots. The study of this pattern showed that the stasis of triaperturate pollen in Eudicots is likely due to stabilizing selection rather than developmental constraints. Experiments on Arabidopsis thaliana mutants also revealed that an increase in aperture number was associated with lower resistance to osmotic stress. This result could suggest that triaperturate pollen grains represent a good trade-off between survival and germination. Moreover, we focused on a particular pollen type with many apertures, to determine its distribution in angiosperms. This type of pollen has many origins, however it is seldom fixed at large taxonomical scale, suggesting the intervention of lineage selection, eliminating this pollen type in the long run.

Magister Philosophiae (Land and Agrarian Studies) - MPhil(LAS) / As a means to assuage historical land inequities, resultant socio-economic disparities and poverty alleviation, the Namibian Government undertook to reform the land sector. Guided by the Constitution and the Resolutions of the 1991 Land Conference policy and legal framework, a fractured consensus is built on the rationale to redistribute land to a targeted group. Parallel to the reform agenda, systemic challenges to the resettlement process are growing amid questions on Government’s ability to respond to sustainable programme objectives embedded within land reforms. Literature coalesces on the issues of land-reform programmes having lost direction, being skewed in favour of a few, being biased towards commercial agriculture, and requiring review and re-configuration to be inclusive and to satisfy equity and povertyalleviation concerns. This study seeks to understand who has been benefitting from land reform by analysing the processes and procedures of identifying beneficiaries prioritised for land allocation, and institutional structures for implementation, while analysing how they produce and reproduce class differentiation and the attendent livelihood trajectories. Using qualitative research conducted in one case study site (Okarukambe constituency) the views of the smallscale farmers who benefitted were solicited. Additionally, the experiences and views of institutions and officials involved in land allocation at regional and national levels are taken into account. Theoretically the study draws from the livelihoods approach to find out the different categories of the small-scale farmers who have benefitted.

No description available.

Purpose: Academic programs are faced with the important task of selecting health professional students who not only possess necessary cognitive abilities to perform their future job, but also have valuable personal/professional characteristics to draw upon in the provision of quality patient care. There is therefore "widespread agreement that it is desirable to broaden the scope of assessment beyond academic achievement" (Eva, Reiter, Trinh, Wasi, Rosenfeld, Norman, 2009, p.768) in candidate selection. The Child Life Studies Program at McMaster University used a 4-station multiple mini-interview (MMI) as part of the admission selection process. This study sought to determine the feasibility, degree of acceptability and reliability of this 4-station MMI design, and if there were any predictors of candidate's performance on the MMI. Methods: A group of 35 applicants in 2014, and 40 applicants in 2015 screened through admission procedures participated in a 4-station MMI. Each station was 15 minutes in length with 5 minutes for scoring each candidate. Anonymous stakeholder surveys were used in 2015 to assess participant and interviewer's perceptions and acceptability of the MMI in applicant selection. Generalizability coefficients were calculated to determine reliability. In addition, candidate's experience in healthcare, professional work experience, experience with children with disabilities, and previous child life course work (or lack thereof) were analyzed through independent t-tests to report any relationship with candidate's performance on the MMI. A one-way ANOVA was also completed to report any relationship between candidate's undergraduate degree type and his/her MMI performance. Results: The 4-station MMI was found to be feasible for a small program with a marked increase found in the number of candidates interviewed in a shorter period of time. It was also found to be highly acceptable among candidates and faculty. However, this studies' findings did not show statistically significant differences in MMI performance based on identified predictors, or undergraduate degree type. The reliability of the 4-station design in 2014 was G=0.718, however, was only G=0.089 in 2015, far lower than expected based on the prior year’s G Coefficient. The potential pitfalls in running a small program MMI are discussed with specific suggestions and modifications provided to enhance reliability of candidate selections across professions. / Thesis / Master of Science (MSc)

The main idea of this thesis is to present the novel concept of localized feature selection (LFS) for data classification and its application for coma outcome prediction. Typical feature selection methods choose an optimal global feature subset that is applied over all regions of the sample space. In contrast, in this study we propose a novel localized feature selection approach whereby each region of the sample space is associated with its own distinct optimized feature set, which may vary both in membership and size across the sample space. This allows the feature set to optimally adapt to local variations in the sample space. An associated localized classification method is also proposed. The proposed LFS method selects a feature subset such that, within a localized region, within-class and between-class distances are respectively minimized and maximized. We first determine the localized region using an iterative procedure based on the distances in the original feature space. This results in a linear programming optimization problem. Then, the second method is formulated as a non-linear joint convex/increasing quasi-convex optimization problem where a logistic function is applied to focus the optimization process on the localized region within the unknown co-ordinate system. This results in a more accurate classification performance at the expense of some sacrifice in computational time. Experimental results on synthetic and real-world data sets demonstrate the effectiveness of the proposed localized approach. Using the LFS idea, we propose a practical machine learning approach for automatic and continuous assessment of event related potentials for detecting the presence of the mismatch negativity component, whose existence has a high correlation with coma awakening. This process enables us to determine prognosis of a coma patient. Experimental results on normal and comatose subjects demonstrate the effectiveness of the proposed method. / Dissertation / Doctor of Philosophy (PhD) / This study proposes a novel form of pattern classification method, which is formulated in a way so that it is easily executable on a computer. Two different versions of the method are developed. These are the LFS (localized feature selection) and lLFS (logistic LFS) methods. Both versions are appropriate for analysis of data with complex distributions, such as datasets that occur in biological signal processing problems. We have shown that the performance of the proposed methods is significantly improved over that of previous methods, on the datasets that were considered in this thesis. The proposed method is applied to the specific problem of determining the prognosis of a coma patient. The viability of the formulation and the effectiveness of the proposed algorithm are demonstrated on several synthetic and real world datasets, including comatose subjects.

No description available.

Mate selection has garnered much attention in the existing literature. However, most mate selection research has reviewed mate selection preferences and criteria individually. In this study, the researcher attempted to illustrate mate selection as an interactive process in which individuals are affected by external influences, and their mate selection criteria are influenced by their self-appraisals and their perceptions of others' mate selection criteria. Two studies were conducted. Study 1 was based on social exchange theory, sexual strategies theory, and social context frameworks, and used multiple-group structural equation modeling to describe the relationships among gender, receptivity to external influences on mate selection, self-perceived relative mate selection position, and relative mate selection demand, between Chinese and American never-married heterosexual adults. The results indicated that the model fit the data well. Self-perceived relative mate selection position and relative mate selection demand were negatively correlated. Women had a higher relative mate selection demand than men did. Self-perceived relative mate selection position fully mediated the effect of receptivity to external influences on relative mate selection, though the indirect effect was not significant. Path values did not differ between Chinese participants and American participants. Study 2 confirmed that the model fit the data well and replicated all significant correlations among latent variables found in Study 1. Additionally, Study 2 found that receptivity to external influences and self-perceived relative mate selection position were positively correlated, and that receptivity to external influences had a negative indirect effect on relative mate selection demand, fully mediated by relative mate selection position. Lastly, the researcher discussed findings, implications, strengths, limitations, and future directions of the present study. / Ph. D.

Forest trees are an important pool of biodiversity at the gene, individual and an ecosystem level. This variation is a result of complex environmental interactions, as well as neutral and selective forces acting on populations. Patterns of standing genetic variation are the result of adaption to past and contemporary climate change, but also historical demographic events, and disentangling the role of these forces is a central problem in population genomics. The overall goal of this study is to characterize the relative effects of demography and selection in the genome of Populus trichocarpa, a riparian deciduous tree species of North America. Specifically, I used a variety of methods to summarize patterns of genetic diversity and population structure in P. trichocarpa, and to reconstruct its demographic history. I subsequently incorporated these demographic insights to guide the application of several methods to identify genome-wide targets of natural selection within and among rangewide populations adapted to heterogeneous selection regimes. Results of this study provide insights into the history of divergence and differentiation in P. trichocarpa populations and help us identify the functional genetic variants contributing to phenotypic divergence and fitness of the individuals in it. / Ph. D.

Digitized by Kansas Correctional Industries