Similarities between golden mice (Ochrotomys nuttalli) and white-footed mice (Peromyscus leucopus) have been well-studied in both field and laboratory settings. Often sympatric, these species share similar habitat, as well as other resources, yet previous researchers have found little evidence for interspecific competition. Niche partitioning may reduce direct competition through specialization of resource use. Although the golden mouse is considered a resource specialist, it is likely that the degree of habitat specialization differs by locality, and thus, the degree of interspecific competition with similar species is variable. To determine the extent to which microhabitat use differs between golden mice and white-footed mice, I measured 16 and 15 microhabitat variables during the leaf-on and leaf-off seasons, respectively, in Jackson County, Illinois. Trapping took place on 3 grids from March 2010 to September 2011. The ratio of individual golden mice (n = 74) to white-footed mice (n = 85) was unusually high during this study. Microhabitat use models were constructed for both species during both seasons using logistic regression by comparing microhabitat at trapping stations where each species was captured vs. stations without captures. Few variables described habitat occupied by golden mice. Overall, dense vegetation up to 2.0 m was most important for golden mice and ground-level structures such as logs were most important for white-footed mice. Captures and noncaptures were predicted with a high degree of accuracy by logistic regression (81.5-90.3%). Discriminant function analysis was used to identify which microhabitat variables optimally discriminated between habitat used by golden mice, white-footed mice, and neither species. More variables discriminated between species during the leaf-on season than the leaf-off season although discriminating variables during leaf-off were more important overall. Habitat where either species was captured was combined and compared against habitat where no mice were captured; microhabitat used by mice was statistically distinct within the study area. Captures and noncaptures were classified correctly more than would be expected by chance by discriminant function analysis but moderate classification success values indicated microhabitat differences between species were subtle. Microhabitat and elevated trap use varied between seasons for both species, but neither species used ground or elevated traps more than expected during the leaf-on season. Spatial segregation was more apparent during the leaf-off season when golden mice used elevated traps more than expected and white-footed mice used ground traps more than expected. Overall results suggest that golden mice exhibited a great deal of plasticity in microhabitat use seasonally, and are more habitat generalists than previous literature would suggest. Although some spatial segregation was apparent between golden mice and white-footed mice, there was no evidence for avoidance between species, which implies a lack of interference competition. It is likely that other life-history factors (such as metabolic rate, nest building, or sociality) in combination with microhabitat and vertical partitioning allow coexistence between these species rather than microhabitat segregation alone.

A chevron is a U-shaped rock structure constructed for improving navigation conditions by diverting majority of flow towards main channel. The objective of this study is to improve understanding of how chevrons affect channel flow. For this study, a two-dimensional numerical hydrodynamic model of a two-km-long reach of the Mississippi River was developed; three chevrons have been constructed in the modeled reach. The model was calibrated by adjusting Manning's n to match predicted and observed water surface elevations (WSELs). The model was validated using measured WSEL and velocity data from two events: a low-flow discharge (4,500 m3/s) and high-flow discharge (14,000 m3/s). At reach scale the model performed well in predicting WSELs. Average difference between model prediction and observed WSEL was 0.23 m in low-flow condition and 0.05 in high flow condition. Root mean square of errors (RMSEs) and mean absolute errors (MAEs) were used to measure the degree of agreement between predicted and measured velocities. At the reach scale there was reasonable agreement between predicted and observed velocities (RMSE = 0.416 m/s and 0.425 m/s, respectively, for low-flow and high-flow conditions). Local differences between predicted and observed velocities were up to 1.5 m/s; this is attributed to uncertainties in the velocity measurements. The model's sensitivity of to changes in Manning's n, eddy viscosity and bathymetry were also analyzed. The sensitivity analysis showed that there are specific areas (e.g., near the banks of the river) which are sensitive to changes in Manning's n. This indicates that spatial distribution of Manning's n is required to increase the accuracy in the model's predictions of velocity. Model was found to be stable in a specific range of eddy viscosity values. Eddy viscosity had little effect on velocity predictions but was important for model stability (i.e., the model was stable only for a range of eddy viscosity values). Reach scale changes in bathymetry had minor impacts on RMSE and MAE. However, local changes in channel bathymetry resulted in differences in velocity predictions as much as ±0.4 m/s.

Representing and handling high-dimensional data has been increasingly ubiquitous in many real world-applications, such as computer vision, machine learning, and data mining. High-dimensional data usually have intrinsic low-dimensional structures, which are suitable for subsequent data processing. As a consequent, it has been a common demand to find low-dimensional data representations in many machine learning and data mining problems. Factorization methods have been impressive in recovering intrinsic low-dimensional structures of the data. When seeking low-dimensional representation of the data, traditional methods mainly face two challenges: 1) how to discover the most variational features/information from the data; 2) how to measure accurate nonlinear relationships of the data. As a solution to these challenges, traditional methods usually make use of a two-step approach by performing feature selection and manifold construction followed by further data processing, which omits the dependence between these learning tasks and produce inaccurate data representation. To resolve these problems, we propose to integrate feature learning and graph learning with factorization model, which allows the goals of learning features, constructing manifold, and seeking new data representation to mutually enhance and lead to powerful data representation capability. Moreover, it has been increasingly common that 2-dimensional (2D) data often have high dimensions of features, where each example of 2D data is a matrix with its elements being features. For such data, traditional data usually convert them to 1-dimensional vectorial data before data processing, which severely damages inherent structures of such data. We propose to directly use 2D data for seeking new representation, which enables the model to preserve inherent 2D structures of the data. We propose to seek projection directions to find the subspaces, in which spatial information is maximumly preserved. Also, manifold and new data representation are learned in these subspaces, such that the manifold are clean and the new representation is discriminative. Consequently, seeking projections, learning manifold and constructing new representation mutually enhance and lead to powerful data representation technique.

While studies of lesbian, gay, and transgender communities and cultural production have dramatically increased, research on bisexuality remains highly undervalued in humanities and social science disciplines. To challenge this lack of scholarship, this doctoral dissertation applies both textual and ethnographic methods to examine bisexual representation in non-realistic or “speculative” narratives and to explore the insider perspectives of bisexual people who are also science fiction fans. The overall trajectory of chapters follows a progression from grounded research and analysis to theory and application. First, I explore bisexual worldviews through ethnographic research in overlapping sexual and fan communities and through textual analysis of a 1980s bisexual fanzine. Next, I establish theoretical and methodological foundations for a new sexual paradigm, called dimensional sexuality, and work to intervene in interpretive methods that may restrict readings of sexuality in cinematic narratives. And finally, I test dimensional sexuality as an interpretive mode by offering dimensional readings of science fiction television and novels. From one direction, the project seeks to understand bisexuality as a position from which to theorize sexual knowledge. A major claim is that bisexual epistemology offers an alternative to dominant monosexual frameworks. Specifically, the multivalent logic of bisexuality refutes the “either-or” structure of heterosexuality and homosexuality. By embracing the logic of “both-and,” bisexuality as a category of knowledge enables the reorganization of sexuality within a non-binary, non-gender based multidimensional framework. From another direction, the project demonstrates the productive textual and social spaces offered by speculative narratives for questioning what we “know” about gender, sex, sexuality, and other intersections of social identities. Science fiction bears a deep structural affinity with the dialectical thinking found in critical theory. By asking “what if” questions that challenge our assumptions about “what is,” non-realistic narratives estrange us from the “known” world, interrogate our assumptions about the world, and make visible ideas and experiences outside of the norms we use to interpret what is “real” in a particular social and historical moment. As such, speculative narratives enable us to imagine sexual and gender possibilities beyond the episteme of the moment.

Submitted by Newton Salvador Grande Neto (newsalgn@hotmail.com) on 2016-04-08T03:13:41Z No. of bitstreams: 1 Newton Salvador corrigido 30-03.pdf: 4627333 bytes, checksum: 32bbc790f76698649a60a184c85c6860 (MD5) / Approved for entry into archive by Ana Paula Grisoto (grisotoana@reitoria.unesp.br) on 2016-04-08T17:05:27Z (GMT) No. of bitstreams: 1 grandeneto_ns_me_ilha.pdf: 4627333 bytes, checksum: 32bbc790f76698649a60a184c85c6860 (MD5) / Made available in DSpace on 2016-04-08T17:05:27Z (GMT). No. of bitstreams: 1 grandeneto_ns_me_ilha.pdf: 4627333 bytes, checksum: 32bbc790f76698649a60a184c85c6860 (MD5) Previous issue date: 2016-03-14 / Coordenação de Aperfeiçoamento de Pessoal de Nível Superior (CAPES) / A técnica de replicar a morfologia de uma estrutura advinda do interior do corpo humano através de um modelo físico é conhecida como biomodelagem. Na área da saúde, um modelo da anatomia humana virtual ou físico é chamado de biomodelo, e este trouxe para a medicina um outro nível em relação a cirurgias modernas, como por exemplo, a possibilidade de o médico cirurgião realizar a simulação de uma cirurgia no biomodelo, analisando as melhores estratégias que serão adotadas para o sucesso da intervenção cirúrgica. Para a confecção de biomodelos são necessárias a execução de três etapas básicas: aquisição de imagens médicas via tomografia computadorizada, tratamento destas imagens utilizando um software específico e a confecção utilizando a manufatura aditiva, caracterizando assim todo o processo de biomodelagem. Todo este processo se tornou possível devido a integração entre as áreas de informática, engenharia, saúde, diagnóstico por imagens e principalmente pelo evento ímpar na área de processos de fabricação, o surgimento da manufatura aditiva. Utilizando um conjunto de tecnologias, a manufatura aditiva é capaz de reproduzir fisicamente, em vários materiais, um modelo virtual camada a camada. Diversas técnicas foram desenvolvidas na área de manufatura aditiva, em especial a impressão tridimensional (3DPrinter) tem seu funcionamento similar a uma impressora comercial a jato de tinta, porém deposita um aglutinante conhecido como binder ao invés de tinta, sobre camadas sucessivas de pó para prototipagem. A reação entre esses dois materiais consolida o formato bidimensional de cada camada, e depois de vários ciclos, um modelo tridimensional está completo. A não utilização de laser para a consolidação das camadas é uma vantagem desta técnica, ou seja, o valor de mercado do maquinário é relativamente mais barato quando comparado a outras técnicas vendidas no mercado. Pesquisas relacionadas a materiais alternativos nacionais são extremamente importantes, pois as descobertas de matérias-primas de baixo custo viabilizam cada vez mais a inclusão da biomodelagem em centros cirúrgicos. Este trabalho teve como objetivo a preparação de um material alternativo economicamente mais viável, utilizando uma proporção em volume de 94% pó de gesso comercial, 5% de ligante e 1% de agente higroscópico. Os resultados demonstram que o material alternativo proposto para este trabalho, se mostrou em torno de 121 vezes mais barato e também atingiu as características necessárias para a construção de biomodelos, como também se mostra tão eficiente em relação a resistência mecânica de manuseio, qualidade superficial e densidade quando comparado a materiais comerciais amplamente aceitos pelo mercado. Com a redução de custos, a técnica de biomodelagem poderá ser utilizada com mais frequência nas intervenções cirúrgicas, diminuindo os riscos existentes na cirurgia através de um planejamento cirúrgico de sucesso. / The technique to replicate a morphology of some interior structure of the human body through a physical model is known as biomodeling. In health care area, a virtual or physical human anatomy model is called biomodel, and this brings to the medicine another level in relation to moderns surgeries, for example, the surgeon has the possibility to perform a simulation of a surgery on a biomodel, making the opportunity to find the best strategies that will be adopted for the success of the surgery intervention. Three basic steps are required to ensure the fabrication of the biomodels: the acquisition of medical images via tomography or MRI, then, the treatment of these images using a specific software, to finally produce the biomodel by additive manufacturing, featuring then the whole process biomodeling. This entire process has become possible because of the integration of information technology, engineering, health, image diagnosis and especially the unique event in the area of manufacturing processes, the emergence of additive manufacturing. By a set of technologies, the additive manufacturing is able to physically reproduce, in several materials, a virtual model layer by layer. Several techniques have been developed in this area, especially the three-dimensional printing (3DPrinter), that operates similarly to a commercial inkjet printer, but, instead of ink, deposits an adhesive known as binder on successive layers of prototyping powder. The reaction between the binder and the powder consolidates the two-dimensional shape of each layer, and, after several cycles, a three-dimensional model is complete. Not utilizing lasers to consolidate the layers is the advantage of this technique that makes the market value of the machinery relatively inexpensive, compared to other market techniques. Researches related to national alternative materials are extremely important, because the Discovery of inexpensive raw materials can enable the inclusion of biomodeling in surgery rooms more and more. The aim of this study is the preparation of an alternative and economically viable material, using a volume proportion of 94% of comercial gypsum powder, 5% of binder and 1% of hygroscopic agent. The results show that the alternative material proposed by this study was about 121 times cheaper and also reached the necessary characteristics for the fabrication of the biomodels, as also shown as efficient regarding to mechanical strength handling, surface quality and density when compared to comercial materials widely accepted by the Market. By reducing the costs, the biomodeling technique can be used more often in surgical interventions, reducing the surgery risks through a success surgical planning.

Master of Science / Department of Statistics / Haiyan Wang / Cancer type classification with high throughput molecular data has received much attention. Many methods have been published in this area. One of them is called PAM (nearest centroid shrunken algorithm), which is simple and efficient. It can give very good prediction accuracy. A problem with PAM is that this method selects too many genes, some of which may have no influence on cancer type. A reason for this phenomenon is that PAM assumes that all genes have identical distribution and give a common threshold parameter for genes selection. This may not hold in reality since expressions from different genes could have very different distributions due to complicated biological process. We propose a new method aimed to improve the ability of PAM to select informative genes. Keeping informative genes while reducing false positive variables can lead to more accurate classification result and help to pinpoint target genes for further studies. To achieve this goal, we introduce variable specific test based on Edgeworth expansion to select informative genes. We apply this test on each gene and select some genes based on the result of the test so that a large number of genes will be excluded. Afterward, soft thresholding with cross-validation can be further applied to decide a common threshold value. Simulation and real application show that our method can reduce the irrelevant information and select the informative genes more precisely. The simulation results give us more insight about where the newly proposed procedure could improve the accuracy, especially when the data set is skewed or unbalanced. The method can be applied to broad molecular data, including, for example, lipidomic data from mass spectrum, copy number data from genomics, eQLT analysis with GWAS data, etc. We expect the proposed method will help life scientists to accelerate discoveries with highthroughput data.

abstract: This thesis presents systematic studies on angle dependent Raman and Photoluminescence (PL) of a new class of layered materials, Transition Metal Trichalcogenides (TMTCs), which are made up of layers possessing anisotropic structure within the van-der-Waals plane. The crystal structure of individual layer of MX3 compounds consists of aligned nanowire like 1D chains running along the b-axis direction. The work focuses on the growth of two members of this family - ZrS3 and TiS3 - through Chemical Vapor Transport Method (CVT), with consequent angle dependent Raman and PL studies which highlight their in-plane optically anisotropic properties. Results highlight that the optical properties of few-layer flakes are highly anisotropic as evidenced by large PL intensity variation with polarization direction (in ZrS3) and an intense variation in Raman intensity with variation in polarization direction (in both ZrS3 and TiS3). Results suggest that light is efficiently absorbed when E-field of the polarized incident excitation laser is polarized along the chain (b-axis). It is greatly attenuated and absorption is reduced when field is polarized perpendicular to the length of 1D-like chains, as wavelength of the exciting light is much longer than the width of each 1D chain. Observed PL variation with respect to the azimuthal flake angle is similar to what has been previously observed in 1D materials like nanowires. However, in TMTCs, since the 1D chains interact with each other, it gives rise to a unique linear dichroism response that falls between 2D and 1D like behavior. These results not only mark the very first demonstration of high PL polarization anisotropy in 2D systems, but also provide a novel insight into how interaction between adjacent 1D-like chains and the 2D nature of each layer influences the overall optical anisotropy of Quasi-1D materials. The presented results are anticipated to have impact in technologies involving polarized detection, near-field imaging, communication systems, and bio-applications relying on the generation and detection of polarized light. / Dissertation/Thesis / Masters Thesis Materials Science and Engineering 2016

Made available in DSpace on 2014-06-11T19:28:33Z (GMT). No. of bitstreams: 0 Previous issue date: 2011-08-19Bitstream added on 2014-06-13T19:58:06Z : No. of bitstreams: 1 claro_gks_me_bauru.pdf: 1638801 bytes, checksum: 64af55b7016b44956f54e04a50b76b3c (MD5) / Coordenação de Aperfeiçoamento de Pessoal de Nível Superior (CAPES) / Este trabalho tem o propósito de contribuir para a generalização tridimensional de problemas de análise numérica da propagação de fissura, mediante a formulação de elementos finitos com descontinuidade incorporada. Em problemas planos as descontinuidades correspondem a linhas que podem ser elaboradas de uma forma relativamente simples, através da construção sequencial de segmentos retos, orientados conforme a direção de falha no interior de cada elemento finito do sólido. Na análise tridimensional a construção do caminho de descontinuidade é mais complexa, pois devem ser construídas superficiais planas no interior de cada elemento e essas superfícies planas devem ser contínuas entre os elementos. É apresentada, nesse trabalho, uma técnica alternativa de construção do caminho de descontinuidade em análises tridimensionais baseado na solução de um problema análogo ao problema de condução de calor, estabelecido a partir de orientações locais de falha, baseado no estado de tensão do problema mecânico. A solução do problema equivalente é obtida utilizando a mesma malha e interpolações do problema mecânico. Para minimizar o esforço computacional, é proposta uma estratégia na qual a análise para mapear o caminho da descontinuidade é restrita ao domínio formado por alguns elementos próximos à superfície de fissura, que se desenvolve ao longo do processo de carregamento. Para validar a metodologia proposta foram realizadas análises tridimensionais de problemas básicos de fratura experimentais e seus resultados foram contrastados com os resultados encontrados na bibliografia. Realizou-se também a comparação do tempo de processamento entre o algoritmo proposto e o algaritmo global para as mesmas análises mencionadas acima. Como resultado, constatou-se que o algoritmo proposto conseguiu descrever satisfatoriamente as trajetórias de descontinuidade, consumindo menor tempo de processamento / This work contributes to the generalization to 3D problems of numerical analysis of crack propagation, through finite elements formulation with embedded discontinuity. In plane problems the discontinuities correspond to lines that can be tracked in a relatively simple way, by sequentially constructing straight segments, following the crack orientation inside each solid finite elements. In tree-dimensional analysis the tracking scheme is more complex since planar surfaces must be constructed inside each element and these planar surfaces must be continuous between elements. It is show in this work, an alternative version of discontinuity path construction technique in three-dimensional analysis based on the solution of an analogous heat conduction problem, established from the local failure orientation based on the stress state of the mechanical problem. The solution of the equivalent problem is obtained using the same mesh and interpolations of the mechanical problem. To minimize computational effort, a strategy is proposed in which the analysis to track the discontinuity path is restricted to the domain formed by few elements near the crack surface front, which develops along the loading process. To validate the poposed methodology three-dimensional analysis of experimental fracture test were performed and the results were contrasted with those obtained from the literature. The comparison between the process time of the proposed algorithm and the global algorithm was performed too. It was found that the proposed algorithm was able to describe the discontinuity path satisfactorily with reduced computational time

Made available in DSpace on 2014-06-11T19:24:10Z (GMT). No. of bitstreams: 0 Previous issue date: 2007-12-12Bitstream added on 2014-06-13T20:51:36Z : No. of bitstreams: 1 spinosa_rmo_me_bauru.pdf: 2012402 bytes, checksum: 866eebe6287f32c6343bf9c863891ba8 (MD5) / Dados antropométricos são os principais parâmetros projetuais para o dimensionamento de produtos e espaços de atividades. Eles permitem predizer alguns aspectos importantes das áreas de interface ainda durante o desenvolvimento projetual, principalmente no que se refere à adequação dimensional para populações específicas. Entretanto, apesar de expressivos, na forma em que são disponibilizados atualmente, estes dados acabam encontrando dificuldades para se difundir, principalmente dentro do setor produtivo onde a velocidade de produção e as exigências do empregador, muitas vezes fazem com que informações essenciais acabem sendo desconsideradas. Outra questão que se levanta refere-se à possível adequação e melhoria no processo de uso desse corpo de conhecimento para o dimensionamento dos projetos atuais, que em sua maioria, caminham para o desenvolvimento tridimensional em ambientes virtuais. Esta pesquisa tem por objetivo propor uma nova abordagem para autilização dos dados antropométricos de populações extremas, através do desenvolvimento de uma interface gráfica de consulta e manipulação, que divulgue e facilite o acesso às medidas, tabulações e parâmetros atualmente disponíveis, fornecendo também, referências e estudos tridimensionais com manequins antropométricos, dimensionalmente confiáveis, da população infantil e de idosos brasileiros, adequando-os desta forma ao contexto tecnológico atual de desenvolvimento de projeto / Data anthropometrics are the principal parameters projects for the dimension of products and spaces of activities. They allow stilling predicting some important aspects of the interface areas during the development projects, mainly in what he/she refers to the dimensional adaptation for specific populations. However, in spite of expressive, in the form in that they are now available, these data end up having difficulties to diffuse, mainly inside of the productive section where the production speed and the demands of the employer, a lot of time do with that essential information end up being inconsiderate. Another subject that one lift refers to the possible adaptation and improvement in the process of use of that knowledge body for the dimension of the current projects, that in your majority, they walk for the three-dimensional development in virtual atmospheres. This research has for objetive to propose a new approach for use of the data anthropometrics of extreme populations, through the development of a graphic interface of consultation and manipulation, that it publishes and facilitate the acess now to the measure, tabulations and parameters available, also supplying, references and three-dimensional studies with dolls anthropometrics, reliable dimension, of the infantile population and of Brazilian senior, adapting them this way to the current technological context of project development

Made available in DSpace on 2014-10-09T12:36:36Z (GMT). No. of bitstreams: 0 / Made available in DSpace on 2014-10-09T13:59:10Z (GMT). No. of bitstreams: 1 04131.pdf: 2671874 bytes, checksum: f1aecab51efb7083cb98abad64e8c2ba (MD5) / Dissertacao (Mestrado) / IPEN/D / Instituto de Pesquisas Energeticas e Nucleares - IPEN/CNEN-SP