Typescript, etc. / Digitized by Kansas State University Libraries

Rhythm is at the heart of all music. It is the variation of the duration of sound over time. A rhythm has two components: one is the striking of an instrument – called the "onset" – and the other is silence. Historically, musical forms and works were preferred and became popular by their rhythmic properties. Therefore, to study rhythm is to study the underpinnings of all of music. In this thesis, we explore basic rhythmic preferences in traditional music and, using this as a point of reference, methods are implemented to generate similar types of rhythms. Finally, a software platform to facilitate such an analysis is developed – it is the first of its kind available to our best knowledge as this research field has only recently emerged.

by Tang Hing Sing. / Thesis (M.Ph.)--Chinese University of Hong Kong, 1989. / Bibliography: leaves 92-96.

Kung Kin Chung. / Thesis (M.Phil.)--Chinese University of Hong Kong, 1997. / Includes bibliographical references (leaves 60-63). / Chapter 1 --- Introduction --- p.1 / Chapter 2 --- 2-D Thermoelastic System --- p.5 / Chapter 2.1 --- Introduction to the thermo-elastic system --- p.5 / Chapter 2.2 --- Existing theoretical results on the system --- p.6 / Chapter 2.2.1 --- Well-posedness and energy decay phenomena of the thermoelastic system --- p.7 / Chapter 2.2.2 --- Reduction of the coupled thermoelastic sys- tem to a single heat equation --- p.8 / Chapter 2.3 --- Some existing numerical methods --- p.12 / Chapter 2.3.1 --- Explicit scheme --- p.12 / Chapter 2.3.2 --- Implicit scheme --- p.16 / Chapter 2.4 --- The formulation and transformation --- p.17 / Chapter 2.5 --- Finite element discretizations --- p.19 / Chapter 2.6 --- Existence of the discrete solution --- p.20 / Chapter 2.7 --- Numerical results --- p.23 / Chapter 2.7.1 --- Introduction to the FEM software --- p.23 / Chapter 2.7.2 --- Numerical experiments --- p.28 / Chapter 3 --- 3-D Elliptic Interface Problem --- p.35 / Chapter 3.1 --- Introduction to 2-D elliptic problems --- p.35 / Chapter 3.2 --- Finite element discretization --- p.36 / Chapter 3.3 --- Convergence of the numerical methods --- p.38 / Chapter 3.4 --- Numerical results --- p.43 / Chapter 3.4.1 --- Introduction --- p.43 / Chapter 3.4.2 --- Introduction to the FEM software --- p.45 / Chapter 3.4.3 --- Numerical experiments --- p.47 / Chapter 4 --- 2-D Parabolic Interface Problem --- p.53 / Chapter 4.1 --- Introduction to the 2-D parabolic problem --- p.53 / Chapter 4.2 --- Finite element discretization --- p.54 / Chapter 4.3 --- Existence and convergence analysis --- p.55 / Chapter 4.4 --- Numerical results --- p.57 / Chapter 4.4.1 --- Introduction --- p.57 / Chapter 4.4.2 --- Numerical experiments --- p.58 / Bibilography

Resumo: Para o aprimoramento de técnicas de diagnóstico e tratamentos de doenças o desenvolvimento de materiais magnéticos é de primordial importância. Neste trabalho as ferritas de manganês e zindo de fórmula geral Mn(1-x)ZNxFe204, onde x assume os valores entre 0,15 e 0,30, foram preparadas pelo método dos precursores poliméricos e caracterizadas por: análises térmicas (termogravimetria associada com análise térmica diferencial (TG/DTA), análise composicional semi-quantitativa por energia dispersiva de raios X (EDX), difratometria de raios X (DRX), refinamento estrutural (Método de Rietveld), adsorção de N2 a 77K (BET), microscopia eletrônica de varredura (MEV), microscopia óptica, espectroscopia de absorção atômica (testes de lixiviação) e medidas de susceptibilidade e permeabilidade magnética. As amostras de ferritas foram também revestidas com quitosana e analisadas com algumas dessas técnicas. Os materiais foram desenvolvidos com o objetivo de utilizá-los como rastreador em biosusceptometria de corrente alternada(BAC), um método de diagnóstico. O método de síntese empregado proporcionou materiais monofásicos em ferrita com baixa porosidade e área superficial. Os materiais apresentaram baixa lixiviação de metais em soluções que simulam os meios fisiológicos do trato gastrointestinal. O revestimento com quitosana foi eficiente suprimindo a dissolução de metais em meios com pHs 7,5 e 6,8. Em pH 1,2, o revestimento se mostrou menos eficiente, porém, ainda assim foi capaz de reduzir a dissolução das espécies que constituem a ferrita. Quanto ao comportamento magnético, as amostras apresentaram magnetizações satisfatórias em 10K e em temperatura ambiente. A amostra contendo zinco a x=0,15 apresentou a mais alta saturação magnética em temperatura ambiente e foi utilizada em um teste com a técnica de BAC apresentando resultados satisfatórios / Abstract: The development of magnetic materials is of primary importance for the improvement of diagnostic techniques and treatments of diseases. In this work, the manganese and zinc ferrites with general formula of Mn(1-x)ZNxFe204 and 0,15< x <0,30, was prepared by polymeric precursor method and characterized by the physical techniques as: thermal analysis [thermogravimetry and differential thermal analysis (TG/DTA)], ssemi-quantitative analysis of the composition by energy dispersive X ray (EDX), X ray diffractometry (XRD), structural refinement (Rietveld Method), N2 adsorption analysis at 77k (BET), scanning electron microscopy (SEM), optical microscopy, atomic absorption spectroscopy (lixiviation test) and magnetic susceptibility measurements. Chitosan coated samples were also prepared which were analyzed by some of those technique. The materials were developed to be used as a tracer in the alternating current biosusceptometry (ACB), a diagnostic method. The employed synthesis method provided monophasic samples of ferrites with low porosity and surface area. The materials showed low metal solubility in the media that simulate gastrointestinal tract solutions. The chitosan coating was efficient to suppress the metal dissolution in solutions with pHs 7.5 and 6.8. In pH=1.2 medium, the coating was not efficient, but still reduced the resolution of ferrite components. The magnetic measurements showed that the samples have satisfactory magnetizations at 10K and room temperature. The sample with x=0.15 exhibited the highest magnetic saturation at room temperature and it was applied in an ACB technique which showed satisfactory results / Orientador: Margarida Juri Saeki / Coorientador: Paulo Noronha Lisboa Filho / Banca: Luis Gallego Martinez / Banca: Andris Figueiroa Bakuzis / O programa de Pós-graduação em Ciência e Tecnologia de Materiais PosMat, tem caráter institucional e integra as atividades de pesquisa em materiais de diversos campi da UNESP / Mestre

No description available.

Markov chain Monte Carlo (MCMC) is one of the most popular statistical inference methods in machine learning. Recent work shows that a significant improvement of the statistical efficiency of MCMC on complex distributions can be achieved by exploiting geometric properties of the target distribution. This is known as geometric MCMC. However, many such methods, like Riemannian manifold Hamiltonian Monte Carlo (RMHMC), are computationally challenging to scale up to high dimensional distributions. The primary goal of this thesis is to develop novel geometric MCMC methods applicable to large-scale problems. To overcome the computational bottleneck of computing second order derivatives in geometric MCMC, I propose an adaptive MCMC algorithm using an efficient approximation based on Limited memory BFGS. I also propose a simplified variant of RMHMC that is able to work effectively on larger scale than the previous methods. Finally, I address an important limitation of geometric MCMC, namely that is only available for continuous distributions. I investigate a relaxation of discrete variables to continuous variables that allows us to apply the geometric methods. This is a new direction of MCMC research which is of potential interest to many applications. The effectiveness of the proposed methods is demonstrated on a wide range of popular models, including generalised linear models, conditional random fields (CRFs), hierarchical models and Boltzmann machines.

No description available.

Bibliography : leaves 85-115

<p>Stochastic volatility models have been focus for research in recent years.</p><p>One interesting and important topic has been the estimation procedure.</p><p>For a given stochastic volatility model this project aims to compare two</p><p>methods of parameter estimation.</p>