"The purposes of this study were: (1) to investigate the effectiveness of the computer-student interactive method in presenting statistical concepts and in instructing students in the applications of these concepts, and (2) to develop instruments that test for the understanding of these concepts and the mastery of these application skills"--Abstract. / Typescript. / "Spring Semester, 1990." / "Submitted to the Department of Curriculum and Instruction in partial fulfillment of the requirements for the degree of Doctor of Philosophy." / Advisor: E. T. Denmark, Professor Directing Dissertation. / Includes bibliographical references.

We have considered the problem in which a biased sample is selected from a finite population, and this finite population itself is a random sample from an infinitely large population, called the superpopulation. The parameters of the superpopulation and the finite population are of interest. There is some information about the selection mechanism in that the selection probabilities are linearly related to the measurements. This is typical of establishment surveys where the selection probabilities are taken to be proportional to the previous year's characteristics. When all the selection probabilities are known, as in our problem, inference about the finite population can be made, but inference about the distribution is not so clear. For continuous measurements, one might assume that the the values are normally distributed, but as a practical issue normality can be tenuous. In such a situation a transformation to normality may be useful, but this transformation will destroy the linearity between the selection probabilities and the values. The purpose of this work is to address this issue. In this light we have constructed two models, an ignorable selection model and a nonignorable selection model. We use the Gibbs sampler and the sample importance re-sampling algorithm to fit the nonignorable selection model. We have emphasized estimation of the finite population parameters, although within this framework other quantities can be estimated easily. We have found that our nonignorable selection model can correct the bias due to unequal selection probabilities, and it provides improved precision over the estimates from the ignorable selection model. In addition, we have described the case in which all the selection probabilities are unknown. This is useful because many agencies (e.g., government) tend to hide these selection probabilities when public-used data are constructed. Also, we have given an extensive theoretical discussion on Poisson sampling, an underlying sampling scheme in our models especially useful in the case in which the selection probabilities are unknown.

Researchers on the Northwest Coast (NWC) are often interested in complex questions regarding social organization, resource intensification, resource control, and impacts of environmental change on resources and in turn human groups. However, the excavation strategies used on the NWC often do not provide the spatial and chronological control within a site that is necessary to document their variability and answer these research questions. The Čḯxwicən site has the potential to address some of the limitations of previous Northwest Coast village site excavations because of its unique and robust sampling strategy, the wide expanse of time that it was occupied, and the multiple house structures present. An on-going project is examining changing human ecodynamics over the breadth of site occupation, focusing on zooarchaeology and geoarchaeological records. This site, located on the Strait of Juan de Fuca in Port Angeles, WA was excavated in 2004 as part of a Washington State Department of Transportation (WSDOT) undertaking to build parts for the Hood Canal Bridge Large scale excavation (261.4 m3 528 m2) generated enormous quantities of faunal remains. Radiocarbon dates and historic records show occupation extends from 2750 cal. BP to the early 20th century. Statistical sampling methods provide an empirical way to maximize the amount of information obtained with the least amount of effort. My thesis addressed the utility of Sampling to Redundancy (STR) as a statistical sampling method for sampling faunal remains from large village sites. My project has documented the variability of fish family representation across time and space in one part of the Čḯxwicən village, while minimizing the time and effort required to do so. This thesis applies STR to "S" (> 1/4 in.) 10 Liter bucket samples from eight excavation units and a total of 26 separate unique temporal and spatial contexts. I focused on 1/4 in. samples for my study for a particular reason. Previous fish faunal studies have focused on effects of mesh size on fish representation; and emphasized the need to use fine mesh (e.g., 1/8 in. or finer) to document small-bodied fishes. This focus on fine mesh typically means that only limited volumes of matrix are studied, which in turn may mean that remains of rarer, large- bodied fishes are under- represented. The on-going research project has focused on buckets screened to 1/8 in. mesh (called "C" buckets). I used STR to sample additional volumes of matrix screened to 1/4 in. to examine whether expanding the volume studied would affect fish representation, which was a second goal of my project. Overall, I studied remains from 269 "S" buckets out of a total of 419 buckets, or 47% of the buckets. STR was most helpful for six of the high bone abundance and density contexts, where I analyzed less than 50% of the total buckets, was moderately helpful for 14 contexts, and not at all helpful for the six contexts with low fishbone abundance, where I analyzed 100% of the buckets. This analysis took me a total of 154 hours, and based on the percentage of material analyzed, 174 hours were saved. As to the second project goal, to assess whether adding fish remains documented from additional matrix volume affected fish representation, I found the differences were minimal. Both for my study units as a whole, and for each time period, adding the fish records from the "S" buckets did not alter the main trends in fish representation as documented by the larger study, using a smaller volume. To further examine whether the added volume from >1/4 in "S" buckets affected results, I explored specific research questions that are relevant to the larger project regarding environment-animal interactions and fishbone deposition and bone condition inside and outside of a house structure. Adding the "S" bucket samples did not affect fish representation or fishbone distribution and condition, which affirms that the sampling strategy used in the larger research project was sufficient in most cases to characterize the fish record at the site. My approach to STR has focused on fish remains that were previously excavated from a Pacific coastal village site with dense archaeological deposits. STR could be employed in other types of archaeological settings in a range of environments (coast or interior) representing a range of cultural contexts (from hunting camps to urban centers) to establish sample redundancy after an excavation is complete. STR could be used during on-going excavation. Further research is required to explore the implications of STR in these settings, however it is likely that the success of STR in other contexts will be dependent on the density and overall abundance of remains, the diversity or material types being studied, as well of course in the range and specificity of questions in each case.

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.

Made available in DSpace on 2015-03-26T13:32:22Z (GMT). No. of bitstreams: 1 texto completo.pdf: 445502 bytes, checksum: 7d042d373f0007ddc1967700e0e477c8 (MD5) Previous issue date: 2014-10-22 / Coordenação de Aperfeiçoamento de Pessoal de Nível Superior / Brucellosis is an infectious and contagious disease caused by bacteria of the genus Bru- ceIIa. It produces a characteristic infection in animals and may aIso contaminate humans. The tests used for diagnosis of bruceIIosis in Brazil are conducted in samples obtained from animals suspected of having the disease, sIaughtered or that died at the farm. In Brazil there are few vaIidation studies of diagnostic tests for bruceIIosis presenting statistical methodologies for the estimation of sensitivity and Specificity satisfactorily. The present work used the methodology proposed by Joseph, Gyorkos e Coupal (1995) to obtain es- timates of sensitivity and Specificity of screening test Buffered Acidified Antigen (AAT) and confirmatory tests Mercaptoethanol (2-ME) and Bacteriological Diagnosis (DBAC), these tests are in accordance with the National Program for Control and Eradication of AnimaI Brucellosis and Tuberculosis (PNCEBT), which began in Brazil in 2001. The study was complemented with comparisons of the estimates in three different scenario: (i) when the resuIt of onIy one test is avaiIabIe (ii) when results of two test are avaiIabIe, and (iii) results of three tests. The data contained samples from 175 animaIs, obtained by convenience from material sent from aII regions of BraziI, to the Laboratory of Diag- nosis of Bacterial Diseases of LANAGRO-MG, between the years 2008-2011. Statistical computations and Gibbs Sampler algorithm were impIemented in OpenBUGS. Results showed that the estimated prevalence of bovine bruceIIosis in suspected animals is 79%, which means that for every 100 animals suspected of having the disease, 79 are diagnosed with it. Regarding performance measures, AAT was more sensitive for diagnosing Bovine BruceIIosis, 2-ME more Specific for not diagnosing Bovine Brucellosis and DBac showed 100% specific for not diagnosing the disease and Iess sensitive for diagnosing the disease. / A brucelose é uma doença infectocontagiosa provocada por bactérias do gênero Brucella que produz infecção característica nos animais, podendo contaminar o homem. Os testes para diagnóstico da brucelose utiIizados no Brasil são reaIizados a partir de amostras obti- das em animais com suspeita da enfermidade abatidos ou mortos na propriedade. No país são poucos os estudos de vaIidação de testes diagnósticos para brucelose que apresentam metodologias estatísticas para a estimação da sensibilidade e da especificidade satisfatori- amente. Neste trabalho, empregou a metodologia proposta por Joseph, Gyorkos e Coupal (1995) para obter estimativas da sensibilidade e da especificidade do teste de triagem Antígeno Acidificado Tamponado (AAT) e dos testes confirmatórios Mercaptoetanol (2- ME) e Diagnóstico Bacteriológico (DBac), testes estes, em conformidade ao Programa Nacional de Controle e Erradicação da Brucelose e da Tuberculose Animal (PNCEBT) inserido no Brasil em 2001. De forma complementar, comparou-se as estimativas em três cenário distintos: quando dispõe do resultado de somente um teste; (ii) quando dispõe dos resultados de dois testes; e (iii) quando dispõe dos resultados de três testes. A amostra conteve 175 animais, obtida por conveniência a partir de material encaminhado, de todas as regiões do BrasiI, ao Laboratório de Diagnóstico de Doenças Bacterianas do LANAGRO-MG, entre os anos de 2008 a 2011. Os códigos para obter as estimativas foram impIementados no OpenBUGS, por meio do algoritmo Gibbs Sampler. Os resuIta- dos apontaram que a prevalência estimada para brucelose bovina em animais suspeitos é de 79%, ou seja, de cada 100 animais com suspeita da doença, 79 são diagnósticos como doentes. Em relação às medidas de desempenho, AAT se mostrou mais sensível para diagnosticar a Brucelose Bovina, o 2-ME mais Especifico para não diagnosticar a Brucelose Bovina e o DBac mostrou-se 100% específico para não diagnosticar a doença e menos sensível para diagnosticar a doença.

Thesis (Ph.D. (Statistics)) -- University of Limpopo, 2019 / This study presents the application of Direct and Indirect methods of Small AreaEstimation(SAE)techniques. Thestudyisaimedatestimatingthetrends and the proportions of households accessing water, sanitation, and electricity for lighting at small areas of the Limpopo Province, South Africa. The study modified Statistics South Africa’s General Household Survey series 2009-2015 and Census 2011 data. The option categories of three variables: Water, Sanitation and Electricity for lighting, were re-coded. Empirical Bayes and Hierarchical Bayes models known as Markov Chain Monte Carlo (MCMC) methods were used to refine estimates in SAS. The Census 2011 data aggregated in ‘Supercross’ was used to validate the results obtained from the models. The SAE methods were applied to account for the census undercoverage counts and rates. It was found that the electricity services were more prioritised than water and sanitation in the Capricorn District of the Limpopo Province. The greatest challenge, however, lies with the poor provision of sanitation services in the country, particularly in the small rural areas. The key point is to suggestpolicyconsiderationstotheSouthAfricangovernmentforfutureequitable provisioning of water, sanitation and electricity services across the country.

Thesis (PhD (Statistics and Actuarial Sciences))--Stellenbosch University, 2008. / We consider the problem of model assessment by risk estimation. Various approaches to risk estimation are considered in a uni ed framework. This a discussion of various complexity dimensions and approaches to obtaining bounds on covering numbers is also presented. The second type of training sample interval estimator discussed in the thesis is Rademacher bounds. These bounds use advanced concentration inequalities, so a chapter discussing such inequalities is provided. Our discussion of Rademacher bounds leads to the presentation of an alternative, slightly stronger, form of the core result used for deriving local Rademacher bounds, by avoiding a few unnecessary relaxations. Next, we turn to a discussion of PAC-Bayesian bounds. Using an approach developed by Olivier Catoni, we develop new PAC-Bayesian bounds based on results underlying Hoe ding's inequality. By utilizing Catoni's concept of \exchangeable priors", these results allowed the extension of a covering number-based result to averaging classi ers, as well as its corresponding algorithm- and data-dependent result. The last contribution of the thesis is the development of a more exible shell decomposition bound: by using Hoe ding's tail inequality rather than Hoe ding's relative entropy inequality, we extended the bound to general loss functions, allowed the use of an arbitrary number of bins, and introduced between-bin and within-bin \priors". Finally, to illustrate the calculation of these bounds, we applied some of them to the UCI spam classi cation problem, using decision trees and boosted stumps. framework is an extension of a decision-theoretic framework proposed by David Haussler. Point and interval estimation based on test samples and training samples is discussed, with interval estimators being classi ed based on the measure of deviation they attempt to bound. The main contribution of this thesis is in the realm of training sample interval estimators, particularly covering number-based and PAC-Bayesian interval estimators. The thesis discusses a number of approaches to obtaining such estimators. The rst type of training sample interval estimator to receive attention is estimators based on classical covering number arguments. A number of these estimators were generalized in various directions. Typical generalizations included: extension of results from misclassi cation loss to other loss functions; extending results to allow arbitrary ghost sample size; extending results to allow arbitrary scale in the relevant covering numbers; and extending results to allow arbitrary choice of in the use of symmetrization lemmas. These extensions were applied to covering number-based estimators for various measures of deviation, as well as for the special cases of misclassi - cation loss estimators, realizable case estimators, and margin bounds. Extended results were also provided for strati cation by (algorithm- and datadependent) complexity of the decision class. In order to facilitate application of these covering number-based bounds,

Estimates of animal abundance or density are fundamental quantities in ecology and conservation, but for many species such as rare, small mammals, obtaining robust estimates is problematic. In this thesis, I combine elements of two standard abundance estimation methods, capture-recapture and distance sampling, to develop a method called trapping point transects (TPT). In TPT, a "detection function", g(r) (i.e. the probability of capturing an animal, given it is r m from a trap when the trap is set) is estimated using a subset of animals whose locations are known prior to traps being set. Generalised linear models are used to estimate the detection function, and the model can be extended to include random effects to allow for heterogeneity in capture probabilities. Standard point transect methods are modified to estimate abundance. Two abundance estimators are available. The first estimator is based on the reciprocal of the expected probability of detecting an animal, ^P, where the expectation is over r; whereas the second estimator is the expectation of the reciprocal of ^P. Performance of the TPT method under various sampling efforts and underlying true detection probabilities of individuals in the population was investigated in a simulation study. When underlying probability of detection was high (g(0) = 0:88) and between-individual variation was small, survey effort could be surprisingly low (c. 510 trap nights) to yield low bias (c. 4%) in the two estimators; but under certain situations, the second estimator can be extremely biased. Uncertainty and relative bias in population estimates increased with decreasing detectability and increasing between-individual variation. Abundance of the Key Largo woodrat (Neotoma floridana smalli), an endangered rodent with a restricted geographic range, was estimated using TPT. The TPT method compared well to other viable methods (capture-recapture and spatially-explicit capture-recapture), in terms of both field practicality and cost. The TPT method may generally be useful in estimating animal abundance in trapping studies and variants of the TPT method are presented.

Sun Yehuan. / "September 2004." / Thesis (Ph.D.)--Chinese University of Hong Kong, 2004. / Includes bibliographical references (p. 213-223). / Electronic reproduction. Hong Kong : Chinese University of Hong Kong, [2012] System requirements: Adobe Acrobat Reader. Available via World Wide Web. / Electronic reproduction. Ann Arbor, MI : ProQuest Information and Learning Company, [200-] System requirements: Adobe Acrobat Reader. Available via World Wide Web. / Mode of access: World Wide Web. / Abstracts in English and Chinese.

The design and evaluation of computer systems rely heavily upon simulation. Simulation is also a major bottleneck in the iterative design process. Applications that may be executed natively on physical systems in a matter of minutes may take weeks or months to simulate. As designs incorporate increasingly higher numbers of processor cores, it is expected the times required to simulate future systems will become an even greater issue. Simulation exhibits a tradeoff between speed and accuracy. By basing experimental procedures upon known statistical methods, the simulation of systems may be dramatically accelerated while retaining reliable methods to estimate error. This thesis focuses on the acceleration of simulation through statistical processes. The first two techniques discussed in this thesis focus on accelerating single-threaded simulation via cluster sampling. Cluster sampling extracts multiple groups of contiguous population elements to form a sample. This thesis introduces techniques to reduce sampling and non-sampling bias components, which must be reduced for sample measurements to be reliable. Non-sampling bias is reduced through the Reverse State Reconstruction algorithm, which removes ineffectual instructions from the skipped instruction stream between simulated clusters. Sampling bias is reduced via the Single Pass Sampling Regimen Design Process, which guides the user towards selected representative sampling regimens. Unfortunately, the extension of cluster sampling to include multi-threaded architectures is non-trivial and raises many interesting challenges. Overcoming these challenges will be discussed. This thesis also introduces thread skew, a useful metric that quantitatively measures the non-sampling bias associated with divergent thread progressions at the beginning of a sampling unit. Finally, the Barrier Interval Simulation method is discussed as a technique to dramatically decrease the simulation times of certain classes of multi-threaded programs. It segments a program into discrete intervals, separated by barriers, which are leveraged to avoid many of the challenges that prevent multi-threaded sampling.