This study focuses on creating a spectral matching algorithm that modifies the real strong ground motions in the time domain by adding wavelets adjustment to the acceleration time series. The spectral matching procedure is at its core a nonlinear problem, thus a nonlinear solving method was employed in the proposed algorithm. The Broyden updating method was selected as the nonlinear solving method because it does not require a differentiation analysis. The Broyden updating also makes use the information of spectral misfit and wavelet magnitudes vector to approximate the Jacobian matrix which expected to give an efficient calculation. A parametric study was numerically conducted to obtain a set of gain factors that reduce the computational time and minimize the spectra misfit. The study was conducted using ten different ground motions, taken from FEMA P-695 (FEMA, 2009), which represent far field, near field-pulse and near field-no pulse earthquake ground motions. A study of compatible wavelet functions was carried out to determine the appropriate wavelet function for the proposed method. The study include the baseline drift, the frequency and time resolution, and the cross correlation between wavelet adjustments during the spectra matching procedure. Based on this study, the corrected tapered cosine wavelet was selected to be used in the proposed algorithm. The proposed algorithm has been tested and compared with other methods that are commonly used in spectral matching; the RSPMatch method and the frequency domain method. The comparing parameters were the computational time, the average misfit, the maximum misfit and error, the PGA, PGV, PGD, the Arias Intensity and the frequency content for both acceleration and displacement time histories. The result showed that the proposed method is able to match the target while preserving the energy development and the frequency content of the original time histories. / Master of Science

Hyperspectral remote sensing image data, such as that recorded by AVIRIS with 224 spectral bands, provides rich information on ground cover types. However, it presents new problems in machine assisted interpretation, mainly in long processing times and the difficulties of class training due to the low ratio of number of training samples to the number of bands. This thesis investigates feasible and efficient feature reduction and image classification techniques which are appropriate for hyperspectral image data. The study is reported in three parts. The first concerns a deterministic approach for hyperspectral data interpretation. Multigroup and multiple threshold spectral coding procedures, and associated techniques for spectral matching and classification, are proposed and tested. By coding on subgroups of bands using one or three thresholds, spectral searching and matching becomes simple, fast and free of the need for radiometric correction. Modifications of existing statistical techniques are proposed in the second part of the investigation A block-based maximum likelihood classification technique is developed. Several subgroups are formed from the complete set of spectral bands in the data, based on the properties of global correlation among the bands. Subgroups which are poorly correlated with each other are treated independently using conventional maximum likelihood classification. Experimental results demonstrate that, when using appropriate subgroup sizes, the new method provides a compromise among classification accuracy, processing time and available training pixels. Furthermore, a segmented, and possibly multi-layer, principal components transformation is proposed as a possible feature reduction technique prior to classification, and for effective colour display. The transformation is performed efficiently on each of the highly correlated subgroups of bands independently. Selected features from each transformed subgroup can be then transformed again to achieve a satisfactory data reduction ratio and to generate the three most significant components for colour display. Classification accuracy is improved and high quality colour image display is achieved in experiments using two AVIRIS data sets.

No description available.

No description available.

Evaluation of the seismic stability of slopes often involves an estimate of the expected sliding displacements. This evaluation requires a suite of acceleration-time histories as input motions. The methods of selecting and scaling these motions can affect the computed sliding displacements. Linear scaling of recorded ground motions and modification of recorded motions by spectral matching are common approaches used for ground motion selection and these approaches were used in this study to select motions for use in sliding displacement analyses. Rigid sliding block analyses and decoupled flexible sliding block analyses were performed using a suite of linearly scaled motions and a suite of spectrally matched motions. . Generally, the spectrally matched motions predict 10 to 30%, on average, smaller displacements and significantly less variability than the linearly scaled motions, when both suites of input motions were developed to match the same acceleration response spectrum. When both suites of input motions were developed to match the same peak ground velocity and acceleration response spectrum, the spectrally matched motions generally predict 5 to 15%, on average, larger displacements than the linearly scaled motions. Because ground motion parameters beyond acceleration response spectrum affect the computed sliding displacement, parameters such as peak ground acceleration (PGA), peak ground velocity (PGV) and mean period (T[subscript m]) should be considered in selecting and scaling motions for use in sliding displacement analyses. / text

This thesis is divided into two topics: the development of a procedure for wind serviceability design of steel buildings and the development of a simple linear response history analysis for building codes. In the United States the building codes are generally silent on the issue of serviceability. This has led to a wide variation in design practices related to service level wind loads. Chapter 2 of this thesis contains a literature review which discusses pertinent aspects of wind drift serviceability, including selecting the mean recurrence interval (MRI), mathematical modeling of the structure, and establishment of rational deformation limits. Chapter 3 contains a journal article submitted to Engineering Journal which describes the recommended procedure for damage based wind serviceability design of steel structures. The procedure uses a broad range of MRIs, bases damage measurement on shear strains, includes all sources of deformation in the model, and bases deformation limits on fragility curves. Chapter 4 of this thesis contains a literature review which examines issues related to performing linear response history analysis. Chapter 5 contains a conference paper submitted to the Tenth U.S. National Conference on Earthquake Engineering which serves as a position paper promoting the inclusion of a linear response history analysis procedure in future editions of the NEHRP Recommended Seismic Provisions and ASCE 7. The procedure address the following issues: selection and scaling of ground motions, the use of spectral matched ground motions, design for dependent actions, and the scaling of responses with the response modification coefficient (R) and the deflection amplification factor (Cd). / Master of Science

Metabolomics aims at the detection and quantitation of metabolites within a biological system. As the most direct representation of phenotypic changes, metabolomics is an important component in system biology research. Recent development on high-resolution, high-accuracy mass spectrometers enables the simultaneous study of hundreds or even thousands of metabolites in one experiment. Liquid chromatography-mass spectrometry (LC-MS) is a commonly used instrument for metabolomic studies due to its high sensitivity and broad coverage of metabolome. However, the identification of metabolites remains a bottle-neck for current metabolomic studies. This thesis focuses on utilizing computational approaches to improve the accuracy and efficiency for metabolite identification in LC-MS/MS-based metabolomic studies. First, an outlier screening approach is developed to identify those LC-MS runs with low analytical quality, so they will not adversely affect the identification of metabolites. The approach is computationally simple but effective, and does not depend on any preprocessing approach. Second, an integrated computational framework is proposed and implemented to improve the accuracy of metabolite identification and prioritize the multiple putative identifications of one peak in LC-MS data. Through the framework, peaks are likely to have the m/z values that can give appropriate putative identifications. And important guidance for the metabolite verification is provided by prioritizing the putative identifications. Third, an MS/MS spectral matching algorithm is proposed based on support vector machine classification. The approach provides an improved retrieval performance in spectral matching, especially in the presence of data heterogeneity due to different instruments or experimental settings used during the MS/MS spectra acquisition. / Master of Science

Descoberto e caracterizado em 1909 por Carlos Chagas, o Trypanosoma cruzi é o agente etiológico da Doença de Chagas. Durante a fase crônica da doença de Chagas, onde a parasitemia é baixa, o diagnóstico se baseia na busca de anticorpos contra os antígenos de T. cruzi no sangue. Para aumentar a certeza do resultado recomenda-se o uso de pelo menos dois métodos sorológicos (geralmente ensaio ELISA, Imunofluorescência Indireta ou Hemaglutinação Indireta) para a confirmação de diagnóstico. Embora os ensaios mencionados sejam de uso amplamente difundido, nenhum deles possui suficiente especificidade para definir o diagnóstico isoladamente, especialmente em pacientes provenientes de regiões onde há sobreposição geográfica com outros parasitos, especialmente do gênero Leishmania ou T. rangeli. Uma alternativa de interesse que foi levantada por alguns autores na década de oitenta é a busca de moléculas de interesse diagnóstico (anticorpos, antígenos ou imunocomplexos) na urina de pacientes. Durante os últimos anos técnicas de proteômica têm sido aplicadas a muitos campos da medicina. Uma das aplicações é na nefrologia para o melhor entendimento da fisiologia renal, explorar a complexidade de mecanismos de doenças e para identificar novos biomarcadores. Além do mais, a maioria das proteínas na urina são glicosiladas e suas propriedades são únicas fazendo delas uma importante fonte de biomarcadores. De maneira geral, percebe-se que apesar dos avanços significativos nos métodos de diagnósticos para a detecção da infecção por T. cruzi, há algumas lacunas que ainda precisam ser preenchidas. O fato de que a sorologia convencional para busca de anticorpos IgGs manter-se-á positiva ao longo da vida do paciente, mesmo após tratamento devido a persistência da resposta imune humoral, é uma limitação, já que não permite ter um critério confiável de cura. Por outro lado, a falta de um sistema rápido e confiável para acompanhar a evolução do tratamento, gera sérias limitações para avaliar o desempenho de novos protocolos de tratamento com fármacos já existentes ou de novos fármacos. Em função dessas limitações, propõe-se desenvolver e validar novos métodos de diagnóstico baseado em espectrometria de massas para a detecção da infecção pelo T. cruzi utilizando a urina como fonte de biomarcadores. / Discovered and characterized in 1909 by Carlos Chagas, Trypanosoma cruzi is the etiologic agent of Chagas\' disease. During the chronic phase of Chagas\' disease, where parasitemia is low, the diagnosis is based on the search for antibodies against T. cruzi antigens in the blood. To increase the certainty of the result, is recommended two serological methods (usually ELISA, Indirect Immunofluorescence or Indirect Hemagglutination) for diagnostic confirmation. Although the aforementioned assays are widely used, none of them has sufficient specificity to define the diagnosis alone, especially in patients from regions where there is geographical overlap with other parasites, especially of the genus Leishmania or T. rangeli. An interesting alternative that was raised by some authors in the 1980s is the use of molecules of diagnostic interest (antibodies, antigens or immunomplexes) in the patients\' urine. During the last years techniques of proteomics have been applied to many fields of medicine. One of the applications is nephrology for the better understanding of renal physiology, to explore the completeness of disease mechanisms and identify new biomarkers. Moreover, most of the proteins in the urine are glycosylated. Their properties are unique, making them an important source of biomarkers. In general, perceived that despite significant advances in diagnostic methods for the detection of T. cruzi infection, there are some gaps that need to be filled. The fact that conventional IgG antibody serology will remain positive over the life of the patient, despite the persistence of the humoral immune response, is a limitation, since it does not allow a reliable criterion of cure. On the other hand, the lack of a reliable system to follow the evolution of the treatment generates serious limitations to evaluate the performance of new treatment protocols with existing drugs or new drugs. Due to these limitations, we propose to develop new methods of diagnosis based on mass spectrometry for the detection of T. cruzi infection using urine as a source of biomarkers.