Thesis (Ph. D.)--Ohio State University, 2005. / Title from first page of PDF file. Includes bibliographical references (p. 115-117).

No description available.

In most classical frameworks for learning from examples, it is assumed that examples are randomly drawn and presented to the learner. In this paper, we consider the possibility of a more active learner who is allowed to choose his/her own examples. Our investigations are carried out in a function approximation setting. In particular, using arguments from optimal recovery (Micchelli and Rivlin, 1976), we develop an adaptive sampling strategy (equivalent to adaptive approximation) for arbitrary approximation schemes. We provide a general formulation of the problem and show how it can be regarded as sequential optimal recovery. We demonstrate the application of this general formulation to two special cases of functions on the real line 1) monotonically increasing functions and 2) functions with bounded derivative. An extensive investigation of the sample complexity of approximating these functions is conducted yielding both theoretical and empirical results on test functions. Our theoretical results (stated insPAC-style), along with the simulations demonstrate the superiority of our active scheme over both passive learning as well as classical optimal recovery. The analysis of active function approximation is conducted in a worst-case setting, in contrast with other Bayesian paradigms obtained from optimal design (Mackay, 1992).

This thesis is concerned with the development of algorithms for simulation optimization (SO), a special case of stochastic optimization where the objective function can only be evaluated through noisy observations from a simulation. Deterministic techniques, when directly applied to simulation optimization problems fail to converge due to their inability to handle randomness thus requiring sophisticated algorithms. However, many existing algorithms dedicated for simulation optimization often show poor performance on implementation as they require extensive parameter tuning. To overcome these shortfalls with existing SO algorithms, we develop ADALINE, a line search based algorithm that eliminates the need for any user defined parameters. ADALINE is designed to identify a local minimum on continuous and integer ordered feasible sets. ADALINE on a continuous feasible set mimics deterministic line search algorithms, while it iterates between a line search and an enumeration procedure on integer ordered feasible sets in its quest to identify a local minimum. ADALINE improves upon many of the existing SO algorithms by determining the sample size adaptively as a trade-off between the error due to estimation and the optimization error, that is, the algorithm expends simulation effort proportional to the quality of the incumbent solution. We also show that ADALINE converges ``almost surely'' to the set of local minima. Finally, our numerical results suggest that ADALINE converges to a local minimum faster, outperforming other advanced SO algorithms that utilize variable sampling strategies. To demonstrate the performance of our algorithm on a practical problem, we apply ADALINE in solving a surgery rescheduling problem. In the rescheduling problem, the objective is to minimize the cost of disruptions to an existing schedule shared between multiple surgical specialties while accommodating semi-urgent surgeries that require expedited intervention. The disruptions to the schedule are determined using a threshold based heuristic and ADALINE identifies the best threshold levels for various surgical specialties that minimizes the expected total cost of disruption. A comparison of the solutions obtained using a Sample Average Approximation (SAA) approach, and ADALINE is provided. We find that the adaptive sampling strategy in ADALINE identifies a better solution quickly than SAA. / Ph. D.

During the design process for an aerospace vehicle, decision-makers must have an accurate understanding of how each choice will affect the vehicle and its performance. This understanding is based on experiments and, increasingly often, computer models. In general, as a computer model captures a greater number of phenomena, its results become more accurate for a broader range of problems. This improved accuracy typically comes at the cost of significantly increased computational expense per analysis. Although rapid analysis tools have been developed that are sufficient for many design efforts, those tools may not be accurate enough for revolutionary concepts subject to grueling flight conditions such as transonic or supersonic flight and extreme angles of attack. At such conditions, the simplifying assumptions of the rapid tools no longer hold. Accurate analysis of such concepts would require models that do not make those simplifying assumptions, with the corresponding increases in computational effort per analysis. As computational costs rise, exploration of the design space can become exceedingly expensive. If this expense cannot be reduced, decision-makers would be forced to choose between a thorough exploration of the design space using inaccurate models, or the analysis of a sparse set of options using accurate models. This problem is exacerbated as the number of free parameters increases, limiting the number of trades that can be investigated in a given time. In the face of limited resources, it can become critically important that only the most useful experiments be performed, which raises multiple questions: how can the most useful experiments be identified, and how can experimental results be used in the most effective manner? This research effort focuses on identifying and applying techniques which could address these questions. The demonstration problem for this effort was the modeling of a reusable booster vehicle, which would be subject to a wide range of flight conditions while returning to its launch site after staging. Contour-based sampling, an adaptive sampling technique, seeks cases that will improve the prediction accuracy of surrogate models for particular ranges of the responses of interest. In the case of the reusable booster, contour-based sampling was used to emphasize configurations with small pitching moments; the broad design space included many configurations which produced uncontrollable aerodynamic moments for at least one flight condition. By emphasizing designs that were likely to trim over the entire trajectory, contour-based sampling improves the predictive accuracy of surrogate models for such designs while minimizing the number of analyses required. The simplified models mentioned above, although less accurate for extreme flight conditions, can still be useful for analyzing performance at more common flight conditions. The simplified models may also offer insight into trends in the response behavior. Data from these simplified models can be combined with more accurate results to produce useful surrogate models with better accuracy than the simplified models but at less cost than if only expensive analyses were used. Of the data fusion techniques evaluated, Ghoreyshi cokriging was found to be the most effective for the problem at hand. Lastly, uncertainty present in the data was found to negatively affect predictive accuracy of surrogate models. Most surrogate modeling techniques neglect uncertainty in the data and treat all cases as deterministic. This is plausible, especially for data produced by computer analyses which are assumed to be perfectly repeatable and thus truly deterministic. However, a number of sources of uncertainty, such as solver iteration or surrogate model prediction accuracy, can introduce noise to the data. If these sources of uncertainty could be captured and incorporated when surrogate models are trained, the resulting surrogate models would be less susceptible to that noise and correspondingly have better predictive accuracy. This was accomplished in the present effort by capturing the uncertainty information via nuggets added to the Kriging model. By combining these techniques, surrogate models could be created which exhibited better predictive accuracy while selecting the most informative experiments possible. This significantly reduced the computational effort expended compared to a more standard approach using space-filling samples and data from a single source. The relative contributions of each technique were identified, and observations were made pertaining to the most effective way to apply the separate and combined methods.

Genome-wide association studies (GWAS) have delivered many novel insights about the etiology of many common heritable diseases. However, in most disorders studied by GWAS, the known single nucleotide polymorphisms (SNPs) associated with the disease do not account for a large portion of the genetic factors underlying the condition. This suggests that many of the undiscovered variants contributing to the risk of common diseases have weak effects or are relatively rare. This thesis introduces novel adaptations of techniques for improving detection power for both of these types of risk variants, and reports the results of analyses applying these methods to real datasets for common diseases. Chapter 2 describes a novel approach to improve the detection of weak-effect risk variants that is based on an adaptive sampling technique known as Distilled Sensing (DS). This procedure entails utilization of a portion of the total sample to exclude from consideration regions of the genome where there is no evidence of genetic association, and then testing for association with a greatly reduced number of variants in the remaining sample. Application of the method to simulated data sets and GWAS data from studies of age-related macular degeneration (AMD) demonstrated that, in many situations, DS can have superior power over traditional meta-analysis techniques to detect weak-effect loci. Chapter 3 describes an innovative pipeline to screen for rare variants in next generation sequencing (NGS) data. Since rare variants, by definition, are likely to be present in only a few individuals even in large samples, efficient methods to screen for rare causal variants are critical for advancing the utility of NGS technology. Application of our approach, which uses family-based data to identify candidate rare variants that could explain aggregation of disease in some pedigrees, resulted in the discovery of novel protein-coding variants linked to increased risk for Alzheimer's disease (AD) in African Americans. The techniques presented in this thesis address different aspects of the "missing heritability" problem and offer efficient approaches to discover novel risk variants, and thereby facilitate development of a more complete picture of genetic risk for common diseases.

Submitted by M?rden L?les (marden.inacio@ufvjm.edu.br) on 2016-07-21T18:35:27Z No. of bitstreams: 2 license_rdf: 0 bytes, checksum: d41d8cd98f00b204e9800998ecf8427e (MD5) Diego_dos_Santos_Vieira.pdf: 3296114 bytes, checksum: b738df6d293269169107da9a2300aaa3 (MD5) / Approved for entry into archive by Rodrigo Martins Cruz (rodrigo.cruz@ufvjm.edu.br) on 2016-07-22T17:07:03Z (GMT) No. of bitstreams: 2 license_rdf: 0 bytes, checksum: d41d8cd98f00b204e9800998ecf8427e (MD5) Diego_dos_Santos_Vieira.pdf: 3296114 bytes, checksum: b738df6d293269169107da9a2300aaa3 (MD5) / Made available in DSpace on 2016-07-22T17:07:03Z (GMT). No. of bitstreams: 2 license_rdf: 0 bytes, checksum: d41d8cd98f00b204e9800998ecf8427e (MD5) Diego_dos_Santos_Vieira.pdf: 3296114 bytes, checksum: b738df6d293269169107da9a2300aaa3 (MD5) Previous issue date: 2015 / Coordena??o de Aperfei?oamento de Pessoal de N?vel Superior (CAPES) / Objetivou-se nesta pesquisa investigar a estrutura e padr?o espacial de uma ?rea de floresta periodicamente inund?vel e suas esp?cies dominantes, de modo a contribuir para defini??o de estrat?gias de manejo e conserva??o (?rea I); analisar a estrutura diam?trica, padr?o espacial e m?todos de amostragem para estimar a densidade de Bertholletia excelsa Bonpl. na Regi?o do Baixo Rio Tapaj?s (?rea II). A ?rea I trata-se de um t?pico fragmento de Floresta Ombr?fila Densa Aluvial utilizado para atender os objetivos do cap?tulo 2, enquanto que a ?rea II representa uma Floresta Ombr?fila Densa de Terra Firme utilizada para atingir os objetivos do cap?tulo 3. Na ?rea I foram instaladas 308 unidades amostrais cont?guas de 10 x 10 m, perfazendo uma ?rea amostral de 3,08 ha. Nessas unidades amostrais, todos os indiv?duos com di?metro a 1,30 m de altura do solo (dap) ? 10 cm foram medidos, identificados e referenciados em coordenadas cartesianas (X,Y). Foram calculados os par?metros fitossociol?gicos de densidade, frequ?ncia, domin?ncia e posi??o sociol?gica, e as esp?cies ordenadas segundo o valor de import?ncia ampliado (VIA). A diversidade e o padr?o espacial foram obtidos atrav?s do ?ndice de Shannon-Weaver e fun??o K de Ripley, respectivamente. Na ?rea II, realizou-se um invent?rio de prospec??o com mapeamento de todas as ?rvores de Bertholletia excelsa Bonpl. com dap ? 20 cm, totalizando 1.000 hectares. Para an?lise da estrutura diam?trica foram utilizadas t?cnicas multivariadas: an?lise de agrupamento e discriminante. Utilizaram-se para an?lise de agrupamento a dist?ncia euclidiana e o m?todo de Ward. O padr?o espacial foi definido por meio do emprego da fun??o univariada K de Ripley. Constatado o padr?o espacial, dividiu-se o mapa em unidades de 50 x 50 m, no qual foram simulados diferentes procedimentos de amostragem (Amostragem Casual Simples, Amostragem Sistem?tica e Amostragem Adaptativa em Cluster) com intensidade amostral de 15% e limite de erro de 10%. As compara??es entre os m?todos foram realizadas por meio do teste F de Graybill e an?lises da precis?o e exatid?o obtidas das 30 simula??es realizadas para cada procedimento. Na ?rea I foram registrados 1.022 indiv?duos, distribu?dos em 64 esp?cies arb?reas e 33 fam?lias. A diversidade registrada foi de 3,03 nats.ind-1. As seis esp?cies de maior VIA foram: Campsiandra laurifolia Beth., Hevea brasiliensis (Willd. ex A.Juss.) Mull.Arg, Glicoxylon pedicellatum Ducke, Tetragastris altissima (Aubl.) Swart, Vantanea parviflora Lam. e Mabea caudata Pax & K.Hoffm. O padr?o espacial da comunidade variou em fun??o da dist?ncia considerada, por?m foi predominantemente agregado. Os padr?es espaciais detectados para as esp?cies foram: predominantemente aleat?rio para a esp?cie Vantanea parviflora; completamente agregado para as esp?cies Hevea brasiliensis e Glicoxylon pedicellatum; predominantemente agregado para Campsiandra laurifolia, Tetragastris altissima e Mabea caudata. Na ?rea II, a distribui??o diam?trica de Bertholletia excelsa apresentou clara tend?ncia ? normalidade, enquanto que o padr?o espacial foi predominantemente aleat?rio. Os m?todos de amostragem adaptativos foram ineficiente e subestimam o n?mero de indiv?duos por hectare. Os procedimentos casual simples e sistem?tico podem ser utilizados para invent?rios florestais de Bertholletia excelsa desde que haja um aumento na intensidade amostral para valores acima de 44% da ?rea total. / Disserta??o (Mestrado) ? Programa de P?s-Gradua??o em Ci?ncia Florestal, Universidade Federal dos Vales do Jequitinhonha e Mucuri, 2015. / ABSTRACT The aim of this research was to investigate the structure and space pattern of a periodically floodable forest area and its prevalent species, in such a way to contribute to the definition of handling and conservation strategies (Area I); analyze the diameter structure, spatial pattern and sampling methods to estimate the Bertholletia excelsa Bonpl density in the Lower Tapaj?s River Region (Area II). Area I is a typical fragment of alluvial Dense Ombrophilous Forest used to accord the objectives of chapter 2, while Area II represent a Solid Ground Dense Ombrophilous Forest used to reach the aims of chapter 3. In Area I 308 contiguous sampling units measuring 10 x10 cm were installed, making a sampling area of 3.08ha. In those sampling units, all the individuals with 1.3m height from the soil (dap) > 10cm were measured, identified and referenced in Cartesian coordinates (X,Y). The phytossociological parameters of density, frequency, dominance and sociological position were calculated, and the species were sorted according to the increased importance rate (VIA). The diversity and the spatial pattern were obtained through the index of Shannon-Weaver and K function of Ripley, respectively. In Area II, an inventory of prospection with the mapping of all Bertholletia excels Bonpl. trees with dap > 20cm was made, totalizing 1,000 hectares. For the analysis of the diametrical structure multi-varied techniques were adopted: grouping analysis and discriminating analysis. To the grouping analysis the Euclidian distance and the Ward method were used. The spatial pattern was defined by employing the univariate Ripley K function. After confirming the spatial pattern, the map was divided into 50 x 50 cm units, in which different sampling procedures were simulated (Simple Casual Sampling, Systematic Sampling and Cluster Adaptive Sampling) with 15% sampling intensity and 10 % error limit. The comparisons between the methods were performed by means of Graybill F test and precision analysis obtained from the 30 simulations performed for each procedure. In Area I 1,022 individuals were recorded, they were distributed into 64 arboreal species and 33 families. The recorded diversity was 3.03 nats.ind-1. The six species with greater VIA were: Campsiandra laurifolia Beth., Hevea brasiliensis (Willd. ex A.Juss.) Mull.Arg, Glicoxyson pedicellatum Ducke, Tetragastris altissima (Aubl.) Swart, Vantanea parviflora Lam. And Mabea caudata Pax & K.Hoffm. The spatial pattern of the community has varied because of the considered distance, however it was predominantly aggregated. The spatial patterns detected for the species were: Predominantly random for the Vantanea parviflora; completely aggregated for the Hevea brasiliensis and Glicoxylon pedicellatum species; predominantly aggregated for Camsiandra laurifolia. Tetragastris altissima and Mabea caudata. In Area II, the diametric distribution of Bertholletia excelsa has showed clear tendency to normality, while the spatial pattern was predominantly random. The adaptive sampling methods were inefficient and underestimate the number of individuals per hectare. The simple casual and systematic procedures can be used in Betholletia excelsa forest inventories if there is an increase of sampling intensity for values over 44% the total area.

Submitted by Jos? Henrique Henrique (jose.neves@ufvjm.edu.br) on 2017-03-08T18:19:10Z No. of bitstreams: 2 license_rdf: 0 bytes, checksum: d41d8cd98f00b204e9800998ecf8427e (MD5) diego_santos_vieira.pdf: 3296114 bytes, checksum: b738df6d293269169107da9a2300aaa3 (MD5) / Approved for entry into archive by Rodrigo Martins Cruz (rodrigo.cruz@ufvjm.edu.br) on 2017-03-30T15:17:05Z (GMT) No. of bitstreams: 2 license_rdf: 0 bytes, checksum: d41d8cd98f00b204e9800998ecf8427e (MD5) diego_santos_vieira.pdf: 3296114 bytes, checksum: b738df6d293269169107da9a2300aaa3 (MD5) / Made available in DSpace on 2017-03-30T15:17:05Z (GMT). No. of bitstreams: 2 license_rdf: 0 bytes, checksum: d41d8cd98f00b204e9800998ecf8427e (MD5) diego_santos_vieira.pdf: 3296114 bytes, checksum: b738df6d293269169107da9a2300aaa3 (MD5) Previous issue date: 2015 / Coordena??o de Aperfei?oamento de Pessoal de N?vel Superior (CAPES) / Objetivou-se nesta pesquisa investigar a estrutura e padr?o espacial de uma ?rea de floresta periodicamente inund?vel e suas esp?cies dominantes, de modo a contribuir para defini??o de estrat?gias de manejo e conserva??o (?rea I); analisar a estrutura diam?trica, padr?o espacial e m?todos de amostragem para estimar a densidade de Bertholletia excelsa Bonpl. na Regi?o do Baixo Rio Tapaj?s (?rea II). A ?rea I trata-se de um t?pico fragmento de Floresta Ombr?fila Densa Aluvial utilizado para atender os objetivos do cap?tulo 2, enquanto que a ?rea II representa uma Floresta Ombr?fila Densa de Terra Firme utilizada para atingir os objetivos do cap?tulo 3. Na ?rea I foram instaladas 308 unidades amostrais cont?guas de 10 x 10 m, perfazendo uma ?rea amostral de 3,08 ha. Nessas unidades amostrais, todos os indiv?duos com di?metro a 1,30 m de altura do solo (dap) ? 10 cm foram medidos, identificados e referenciados em coordenadas cartesianas (X,Y). Foram calculados os par?metros fitossociol?gicos de densidade, frequ?ncia, domin?ncia e posi??o sociol?gica, e as esp?cies ordenadas segundo o valor de import?ncia ampliado (VIA). A diversidade e o padr?o espacial foram obtidos atrav?s do ?ndice de Shannon-Weaver e fun??o K de Ripley, respectivamente. Na ?rea II, realizou-se um invent?rio de prospec??o com mapeamento de todas as ?rvores de Bertholletia excelsa Bonpl. com dap ? 20 cm, totalizando 1.000 hectares. Para an?lise da estrutura diam?trica foram utilizadas t?cnicas multivariadas: an?lise de agrupamento e discriminante. Utilizaram-se para an?lise de agrupamento a dist?ncia euclidiana e o m?todo de Ward. O padr?o espacial foi definido por meio do emprego da fun??o univariada K de Ripley. Constatado o padr?o espacial, dividiu-se o mapa em unidades de 50 x 50 m, no qual foram simulados diferentes procedimentos de amostragem (Amostragem Casual Simples, Amostragem Sistem?tica e Amostragem Adaptativa em Cluster) com intensidade amostral de 15% e limite de erro de 10%. As compara??es entre os m?todos foram realizadas por meio do teste F de Graybill e an?lises da precis?o e exatid?o obtidas das 30 simula??es realizadas para cada procedimento. Na ?rea I foram registrados 1.022 indiv?duos, distribu?dos em 64 esp?cies arb?reas e 33 fam?lias. A diversidade registrada foi de 3,03 nats.ind-1. As seis esp?cies de maior VIA foram: Campsiandra laurifolia Beth., Hevea brasiliensis (Willd. ex A.Juss.) Mull.Arg, Glicoxylon pedicellatum Ducke, Tetragastris altissima (Aubl.) Swart, Vantanea parviflora Lam. e Mabea caudata Pax & K.Hoffm. O padr?o espacial da comunidade variou em fun??o da dist?ncia considerada, por?m foi predominantemente agregado. Os padr?es espaciais detectados para as esp?cies foram: predominantemente aleat?rio para a esp?cie Vantanea parviflora; completamente agregado para as esp?cies Hevea brasiliensis e Glicoxylon pedicellatum; predominantemente agregado para Campsiandra laurifolia, Tetragastris altissima e Mabea caudata. Na ?rea II, a distribui??o diam?trica de Bertholletia excelsa apresentou clara tend?ncia ? normalidade, enquanto que o padr?o espacial foi predominantemente aleat?rio. Os m?todos de amostragem adaptativos foram ineficiente e subestimam o n?mero de indiv?duos por hectare. Os procedimentos casual simples e sistem?tico podem ser utilizados para invent?rios florestais de Bertholletia excelsa desde que haja um aumento na intensidade amostral para valores acima de 44% da ?rea total. / Disserta??o (Mestrado) ? Programa de P?s-Gradua??o em Ci?ncia Florestal, Universidade Federal dos Vales do Jequitinhonha e Mucuri, 2015. / The aim of this research was to investigate the structure and space pattern of a periodically floodable forest area and its prevalent species, in such a way to contribute to the definition of handling and conservation strategies (Area I); analyze the diameter structure, spatial pattern and sampling methods to estimate the Bertholletia excelsa Bonpl density in the Lower Tapaj?s River Region (Area II). Area I is a typical fragment of alluvial Dense Ombrophilous Forest used to accord the objectives of chapter 2, while Area II represent a Solid Ground Dense Ombrophilous Forest used to reach the aims of chapter 3. In Area I 308 contiguous sampling units measuring 10 x10 cm were installed, making a sampling area of 3.08ha. In those sampling units, all the individuals with 1.3m height from the soil (dap) > 10cm were measured, identified and referenced in Cartesian coordinates (X,Y). The phytossociological parameters of density, frequency, dominance and sociological position were calculated, and the species were sorted according to the increased importance rate (VIA). The diversity and the spatial pattern were obtained through the index of Shannon-Weaver and K function of Ripley, respectively. In Area II, an inventory of prospection with the mapping of all Bertholletia excels Bonpl. trees with dap > 20cm was made, totalizing 1,000 hectares. For the analysis of the diametrical structure multi-varied techniques were adopted: grouping analysis and discriminating analysis. To the grouping analysis the Euclidian distance and the Ward method were used. The spatial pattern was defined by employing the univariate Ripley K function. After confirming the spatial pattern, the map was divided into 50 x 50 cm units, in which different sampling procedures were simulated (Simple Casual Sampling, Systematic Sampling and Cluster Adaptive Sampling) with 15% sampling intensity and 10 % error limit. The comparisons between the methods were performed by means of Graybill F test and precision analysis obtained from the 30 simulations performed for each procedure. In Area I 1,022 individuals were recorded, they were distributed into 64 arboreal species and 33 families. The recorded diversity was 3.03 nats.ind-1. The six species with greater VIA were: Campsiandra laurifolia Beth., Hevea brasiliensis (Willd. ex A.Juss.) Mull.Arg, Glicoxyson pedicellatum Ducke, Tetragastris altissima (Aubl.) Swart, Vantanea parviflora Lam. And Mabea caudata Pax & K.Hoffm. The spatial pattern of the community has varied because of the considered distance, however it was predominantly aggregated. The spatial patterns detected for the species were: Predominantly random for the Vantanea parviflora; completely aggregated for the Hevea brasiliensis and Glicoxylon pedicellatum species; predominantly aggregated for Camsiandra laurifolia. Tetragastris altissima and Mabea caudata. In Area II, the diametric distribution of Bertholletia excelsa has showed clear tendency to normality, while the spatial pattern was predominantly random. The adaptive sampling methods were inefficient and underestimate the number of individuals per hectare. The simple casual and systematic procedures can be used in Betholletia excelsa forest inventories if there is an increase of sampling intensity for values over 44% the total area.

CoordenaÃÃo de AperfeiÃoamento de NÃvel Superior / Este trabalho propÃe melhorias em uma tÃcnica de amostragem adaptativa multidimensional para renderizaÃÃo. RenderizaÃÃo à o processo de sÃntese de imagens por meio de algoritmos que simulam a iluminaÃÃo em cenÃrios virtuais. As tÃcnicas mais gerais de renderizaÃÃo fotorrealÃstica â aquelas que procuram obter imagens que se assemelham a fotografias â utilizam mÃtodos de integraÃÃo baseados em Monte Carlo para resolver a equaÃÃo que descreve a distribuiÃÃo de luz na cena (equaÃÃo de renderizaÃÃo). Por ser um mÃtodo probabilÃstico e utilizar amostras geradas randomicamente, Monte Carlo produz ruÃdo na imagem final â resultado da variÃncia das amostras â e portanto, pode necessitar de uma grande quantidade de amostras para que o ruÃdo diminua a nÃveis aceitÃveis. Com o intuito de se obter imagens de melhor qualidade com uma menor quantidade de amostras, foram pospostas tÃcnicas de amostragem adaptativa que visam concentrar o esforÃo de amostragem em regiÃes mais importantes da cena. Neste trabalho, propÃe-se a modificaÃÃo de uma tÃcnica de amostragem adaptativa multidimensional por meio da adiÃÃo de duas etapas: substituiÃÃo de amostras e integraÃÃo auxiliar. Essas etapas visam dar mais robustez à tÃcnica, possibilitando sua utilizaÃÃo em uma maior variedade de situaÃÃes. AlÃm da adiÃÃo de duas etapas, tambÃm propÃe-se uma tÃcnica de reconstruÃÃo mais eficiente na etapa final. / This work proposes improvements in a multidimensional adaptive sampling technique for rendering. Rendering is the process of synthesizing images by algorithms simulating lighting in virtual scenes. The more general techniques of photorealistic rendering â those seeking images that resemble photographs â use integration methods based on Monte Carlo to solve the equation that describes the distribution of light in the scene (rendering equation). Being a probabilistic method which uses randomly generated samples, Monte Carlo produces noise in the final image â result of samplesâ variance â and therefore may require a large amount of samples to reduce the noise to acceptable levels. To obtain images of better quality with a lower number of samples, adaptive sampling techniques were proposed, concentrating sampling effort in the most important regions. In this work, we propose the addition of two steps to a multidimensional adaptive sampling technique: substitution of samples and auxiliary integration. These steps aim to give more strength to the technique, enabling their use in a wider variety of situations.

Proteins and DNA are large, complex molecules that carry out biological functions essential to all life. Their successful operation relies on adopting specific structures, stabilized by intra-molecular interactions between atoms. The spatial and temporal resolution required to study the mechanics of these molecules in full detail can only be obtained using computer simulations of molecular models. In a molecular dynamics simulation, a trajectory of the system is generated, which allows mapping out the states and dynamics of the molecule. However, the time and length scales characteristic of biological events are many orders of magnitude larger than the resolution needed to accurately describe the microscopic processes of the atoms. To overcome this problem, sampling methods have been developed that enhance the occurrence of rare but important events, which improves the statistics of simulation data. This thesis summarizes my work on developing the AWH method, an algorithm that adaptively optimizes sampling toward a target function and simultaneously finds and assigns probabilities to states of the simulated system. I have adapted AWH for use in molecular dynamics simulations. In doing so, I investigated the convergence of the method as a function of its input parameters and improved the robustness of the method. I have also worked on a generally applicable approach for calculating the target function in an automatic and non-arbitrary way. Traditionally, the target is set in an ad hoc way, while now sampling can be improved by 50% or more without extra effort. I have also used AWH to improve sampling in two biologically relevant applications. In one paper, we study the opening of a DNA base pair, which due to the stability of the DNA double helix only very rarely occurs spontaneously. We show that the probability of opening depends on both nearest-neighbor and longer-range sequence effect and furthermore structurally characterize the open states. In the second application the permeability and ammonia selectivity of the membrane protein aquaporin is investigated and we show that these functions are sensitive to specific mutations. / <p>QC 20171117</p>