Thesis (Ph. D.)--University of Chicago, Graduate School of Business. / Includes bibliographical references. Also available on the Internet.

Reductions in the cost and size of integrated circuits are allowing more and more complex functions to be included in previously simple tools such as lawn-mowers, ovens, and thermostats. Because of this, the process of synthesizing such functions from their initial representation to an optimal VLSI implementation is rarely hand-performed; instead, automated synthesis and optimization tools are a necessity. The factors such tools must take into account are numerous, including area (size), power consumption, and timing factors, to name just a few. Existing tools have traditionally focused upon optimization of two-level representations. However, new technologies such as Field Programmable Gate Arrays (FPGAs) have generated additional interest in three-level representations and structures such as Kronecker Decision Diagrams (KDDs). The reason for this is that when implementing a circuit on an FPGA, the cost of implementing exclusive-or logic is no more than that of traditional AND or OR gates. This dissertation investigates the use of the autocorrelation coefficients in logic synthesis for these types of structures; specifically, whether it is possible to pre-process a function to produce a subset of its autocorrelation coefficients and make use of this information in the choice of a three-level decomposition or of decomposition types within a KDD. This research began as a general investigation into the properties of autocorrelation coefficients of switching functions. Much work has centered around the use of a function's spectral coefficients in logic synthesis; however, very little work has used a function's autocorrelation coefficients. Their use has been investigated in the areas of testing, optimization for Programmable Logic Arrays (PLAs), identification of types of complexity measures, and in various DD-related applications, but in a limited manner. This has likely been due to the complexity in their computation. In order to investigate the uses of these coefficients, a fast computation technique was required, as well as knowledge of their basic properties. Both areas are detailed as part of this work, which demonstrates that it is feasible to quickly compute the autocorrelation coefficients. With these investigations as a foundation we further apply the autocorrelation coefficients to the development of a classification technique. The autocorrelation classes are similar to the spectral classes, but provide significantly different information. The dissertation demonstrates that some of this information highlighted by the autocorrelation classes may allow for the identification of exclusive-or logic within the function or classes of functions. In relation to this, a major contribution of this work involves the design and implementation of algorithms based on these results. The first of these algorithms is used to identify three-level decompositions for functions, and the second to determine decomposition type lists for KDD-representations. Each of these implementations compares well with existing tools, requiring on average less than one second to complete, and performing as well as the existing tools about 70% of the time. / Graduate

Made available in DSpace on 2014-06-11T19:35:41Z (GMT). No. of bitstreams: 0 Previous issue date: 2008-08-27Bitstream added on 2014-06-13T19:25:05Z : No. of bitstreams: 1 claro_fae_dr_guara.pdf: 832858 bytes, checksum: c3de712a6a76ca96ccda8bf5752d87f4 (MD5) / Coordenação de Aperfeiçoamento de Pessoal de Nível Superior (CAPES) / Os gráficos de X são apresentados na literatura supondo quase sempre que as observações da variável X são independentes. Na prática, no entanto, está se tornando rotina descobrir que esta condição não existe. A dependência entre observações gera um aumento na freqüência de alarmes falsos e diminui o poder do dispositivo estatístico. Nesta tese estuda-se o gráfico de X com amostragem dupla (AD) supondo que as observações de X são descritas por modelos parcimoniosos da família ARIMA (Autoregressivo, Integrado e de Médias Móveis). As propriedades da carta foram obtidas considerando o conceito de subgrupos racionais. Para comparar o desempenho do gráfico proposto com o desempenho dos esquemas concorrentes, isto é, o gráfico de X padrão, o gráfico de X com amostra de tamanho variável (ATV) e o esquema da Média Móvel Ponderada Exponencialmente (EWMA), foi necessário obter o número médio de amostras até o sinal (NMA) para todos eles. Os resultados obtidos mostram que a autocorrelação dentro do subgrupo tem forte impacto sobre as propriedades dos gráficos. O gráfico de controle com amostragem dupla é geralmente mais eficiente do que os esquemas concorrentes na detecção de desajustes na média do processo. / The X charts are presented in the literature often assuming that the observations of the X variable are independent. In practice, however, it is becoming a routine to find out that such condition is unrealistic. The autocorrelation among the observations increases the false alarm rate and reduces the power of the statistical device. In this thesis, we study the Double Sampling X chart (DS) assuming that the observations of X are described by parsimonious models of the ARIMA family (Autoregressive, Integrated and Moving Average). The properties of the charts were obtained considering the concept of rational subgroups. To compare the performance of the proposed chart with the performance of the competitor schemes, that is, the standard X chart, the Variable Sample Size X chart (VSS) and the Exponentially Weighted Moving Average (EWMA) chart, it was necessary to obtain the average run length (ARL) for all of them. The results show that the autocorrelation within the subgroup has strong impact on the chart properties. The Double Sampling X chart is usually more efficient than the competitor schemes in the detection of the process mean shifts.

This dissertation shows that the autocorrelation of switching functions can be effectively utilized in combinational logic optimization and synthesis. The procedures developed exploit information contained in the autocorrelation of switching functions to perform optimization of Programmable Logic Arrays (PLAs) and to aid in a multi-level logic synthesis approach called two-place decomposition. A new optimization technique is presented, based on the autocorrelation of switching functions, to find near-optimal variable pairings for decoded PLAs. The results of this approach compare favourably to those of other researchers’ techniques. The key advantages of the new approach are its simplicity and its efficiency. The basic two-place decomposition approach is augmented with various enhancements. These include an improved decomposition merge procedure, the addition of alternate mapping functions for complex disjunctive decompositions, and the incorporation of linearization using the autocorrelation to handle functions that are non-two-place decomposable. A robust implementation of the enhanced method is presented and is used to generate function realizations for comparison with other synthesis methods. The enhanced two-place decomposition method is shown to perform particularly well for functions exhibiting high degrees of symmetry. The dissertation also presents a new synthesis technique that utilizes a particular representation of a switching function called a Reduced Ordered Binary Decision Diagram (ROBDD) and is targeted to two-place decomposition. This new technique allows the two-place decomposition approach to synthesize a much broader range of functions. Although, in comparison to one other synthesis method, the new approach does not perform as well in most cases, it has considerable promise and several enhancements are proposed for improvement. This dissertation also shows that there is a strong connection among autocorrelation, two-place decomposition, and good variable orders in an ROBDD. A first attempt to formally analyze the relationship between autocorrelation and two-place decomposition is presented. Relationships are identified between certain autocorrelation coefficients when particular two-place decompositions exist in a function. These relationships are also connected to the heuristics used in the above mentioned PLA optimization technique. Variable order can have a substantial impact on the size of an ROBDD. This dissertation shows that a good variable order is related to the two-place decompositions that are exhibited in a function. Thus, variable order is also related to the autocorrelation and this relationship can lead to an autocorrelation-based technique for determining good variable orders for ROBDDs. / Graduate

Orientador: Antonio Fernando Branco Costa / Banca: Messias Borges Silva / Banca: Linda Lee Ho / Banca: Anderson Paulo de Paiva / Banca: Carlos Alberto Chaves / Resumo: Os gráficos de X são apresentados na literatura supondo quase sempre que as observações da variável X são independentes. Na prática, no entanto, está se tornando rotina descobrir que esta condição não existe. A dependência entre observações gera um aumento na freqüência de alarmes falsos e diminui o poder do dispositivo estatístico. Nesta tese estuda-se o gráfico de X com amostragem dupla (AD) supondo que as observações de X são descritas por modelos parcimoniosos da família ARIMA (Autoregressivo, Integrado e de Médias Móveis). As propriedades da carta foram obtidas considerando o conceito de subgrupos racionais. Para comparar o desempenho do gráfico proposto com o desempenho dos esquemas concorrentes, isto é, o gráfico de X padrão, o gráfico de X com amostra de tamanho variável (ATV) e o esquema da Média Móvel Ponderada Exponencialmente (EWMA), foi necessário obter o número médio de amostras até o sinal (NMA) para todos eles. Os resultados obtidos mostram que a autocorrelação dentro do subgrupo tem forte impacto sobre as propriedades dos gráficos. O gráfico de controle com amostragem dupla é geralmente mais eficiente do que os esquemas concorrentes na detecção de desajustes na média do processo. / Abstract: The X charts are presented in the literature often assuming that the observations of the X variable are independent. In practice, however, it is becoming a routine to find out that such condition is unrealistic. The autocorrelation among the observations increases the false alarm rate and reduces the power of the statistical device. In this thesis, we study the Double Sampling X chart (DS) assuming that the observations of X are described by parsimonious models of the ARIMA family (Autoregressive, Integrated and Moving Average). The properties of the charts were obtained considering the concept of rational subgroups. To compare the performance of the proposed chart with the performance of the competitor schemes, that is, the standard X chart, the Variable Sample Size X chart (VSS) and the Exponentially Weighted Moving Average (EWMA) chart, it was necessary to obtain the average run length (ARL) for all of them. The results show that the autocorrelation within the subgroup has strong impact on the chart properties. The Double Sampling X chart is usually more efficient than the competitor schemes in the detection of the process mean shifts. / Doutor

O presente trabalho abordou o sistema de Tomografia por Coerência Óptica (OCT) e sua aplicação à área de microfluídica. Para tanto, foram realizados testes de caracterização física de circuitos microfluídicos, utilizando modelos 3D (tridimensionais) construídos a partir de imagens de OCT destes circuitos. A técnica foi, assim, avaliada como potencial ferramenta de auxílio na aferição de microcanais. Indo além, este trabalho estuda e desenvolve técnicas de análise para fluxos microfluídicos, em especial técnicas baseadas no padrão de speckle. Em primeiro momento, métodos já existentes foram estudados e aprimorados, como o Speckle Variance OCT, em que foi obtido um ganho de 31% em tempo de processamento. Outros métodos, como o LASCA (Laser speckle Contrast Analysis), baseados na autocorrelação de speckle, são adaptados às imagens de OCT. Derivado do LASCA, o método de análise desenvolvido baseado na autocorrelação de intensidade motivou o desenvolvimento de um arranjo de OCT próprio e software de aquisição customizado, com taxa de amostragem da ordem de 8 kHz. O método proposto foi, então, capaz de distinguir fluxos volumétricos variados, e seus limites de detecção foram testados, comprovando sua viabilidade de aplicação para análise de movimento browniano e fluxos volumétricos abaixo de 10 μl/min. / This work discusses the Optical Coherence Tomography system (OCT) and its application to the microfluidics area. To this end, physical characterization of microfluidic circuits were performed using 3D (three-dimensional) models constructed from OCT images of such circuits. The technique was thus evaluated as a potential tool to aid in the inspection of microchannels. Going further, this workpaper studies and develops analytical techniques for microfluidic flow, in particular techniques based on speckle pattern. In the first instance, existing methods were studied and improved, such as Speckle Variance - OCT, where a gain of 31% was obtained in processing time. Other methods, such as LASCA (Laser Speckle Contrast Analysis), based on speckle autocorrelation, are adapted to OCT images. Derived from LASCA, the developed analysis technique based on intensity autocorrelation motivated the development of a custom OCT system as well as an optimized acquisition software, with a sampling rate of 8 kHz. The proposed method was, then, able to distinguish different flow rates, and limits of detection were tested, proving its feasibility for implementation on Brownian motion analysis and flow rates below 10 μl/min.

Aim It is commonly asserted that climate presents the primary constraint on species’ geographic distributions, and therefore, that species' ranges shift in response to changing climate given their specific climatic tolerances. However, supporting evidence is surprisingly inconsistent. Alternatively, spatially structured processes (e.g., dispersal) could more strongly determine species’ geographic distributions. Is climate the primary determinant of species’ geographic distributions, or might non-climatic, spatial processes constitute a stronger influence, such that the effect of climate is indirect? This study tests a number of predictions made by each of these hypotheses, during a single period of time. Location Contiguous United States and southern Canada. Methods We used 19 species of passerine birds whose distributions fall entirely within the area sampled by the North American Breeding Bird Survey from 1990-2000. We related these distributions to the mean breeding season climate, geographic locations and neighbourhood effects. Two spatial scales were addressed to assess the geographic location of species’ ranges and species' distributions within ranges. Results On average, geographic coordinates and a model representing neighbourhood occupancy outperform a simple climatic model. After controlling for geographic coordinates, species occupancy is poorly related to climate. A neighbourhood model on average accounts for the majority of variance captured by geographic coordinates within ranges, and more for the continental placement of ranges. Spatially explicit variables are more important than macroclimatic variables in a predictive model of species occupancy on average. Main Conclusions The geographic distributions of wide-spread North American passerine birds appear not to be primarily determined by climate. Our results are consistent with the hypothesis that localized spatial processes such as dispersal are stronger determinants of both continental range placement and within-range distributions of North American birds.

No description available.

Asset pricing models such as the CAPM calls for the estimation of beta as a measure of the systematic risk. Using historical betas as an input to portfolio analysis requires the assumption of beta stationarity. The existing literature on beta dynamics suggest a somewhat high dispersion of the beta persistence across stocks. In previously unexplored territory, this study aims to investigate factors associated with the degree of beta persistence. By using a sample of 237 U.S. stocks with daily returns observed over the period 1984 to 2015, yearly stock betas were estimated using a GARCH / Maximum Likelihood framework. Autocorrelation properties of these beta series was then crosssectionally regressed on five hypothesized determining variables. Product type as well as the absolute value of beta was found to have a significant effect on the first-order autocorrelation of beta.

No description available.