In traditional quality control charts, fixed sampling interval (FSI) schemes are used where the time between samples has fixed intervals. More efficient methods called variable sampling interval (VSI) schemes have been developed where one takes the next observation sooner than usual if there is an indication that the process is operating off the target value. Another traditional assumption behind most statistical process control charts is that the sequential observations are independent. However, there are many situations where the sequential observations should not to be treated as independent. Rather, a time series model, in particular the first order autoregressive (AR (1)) model, is appropriate. A Markov chain representation is used to study the properties of the FSI and VSI Shewhart X control charts. First, the results show that if the process variance is properly estimated and if traditional control limits are used in the FSI control charts, then the detection time is shorter when the consecutive observations are negatively correlated than when they are positively correlated. If they are positively correlated, then the false alarm rate decreases as the correlation between consecutive observations increases. On the other hand, the detection time increases as the correlation increases. In VSI control charts with traditional control limits, if the process mean is on or near the target, then the average time to signal (A TS) and average number of samples to signal (ANSS) tend to decrease as the correlation increases until the correlation becomes rather moderate. Then, for more highly correlated data, the A TS and ANSS tend to increase as the correlation increases. Next, the results show that, even under the AR (1) process, the VSI chart is more efficient than the FSI chart in terms of ATS. In contrast, the VSI chart is less efficient than the FSI chart in terms of ANSS. The efficiency (inefficiency) of ATS (ANSS) tends to decrease (increase) as the correlation between the consecutive observations becomes stronger. Steady state ATS (A TS·) and steady state ANSS (ANSSO) under the AR (1) process show the same trend as the 'regular' ATS and 'regular' ANSS except when the deviation is very large. If the deviation is very large, then the VSI control chart does not seem to be more efficient than the FSI control chart in terms of steady state ATS. If we have an AR (2) process, then for any given value of tP2 a PSI control chart has a shorter detection time when tPl is negative than when tPl is positive. In a FSI control chart, the effect of positive </>2 in addition to positive tPl is that the false alarm rate decreases even further and the detection time is even longer. / Ph. D.

M.S.

A dissertation submitted to the faculty of Engineering, University of the Witwatersrand, Johannesburg, in fulfilment of the requirements for the degree of Master of science in Engineering Johannesburg 1991 / This dissertation discusses the interaction between shipboard antennas and their environment. The emphasis is on the Use of the Method of Moments to calculate the currents on the structure of the ship. These currents are induced by the antennas mounted on the structure of the ship. The parameters (such as grid spacing and wire radius) to use in creating a wire grid model of the ship is investigated and recommended values given. A sample ship is analyzed and the results obtained compared with measurements done in an unechoic chamber.

This thesis provides statistical analysis methods and a validation procedure for conducting this statistical analysis, under the common random number (CRN) correlation-induction strategy. The proposed statistical analysis provides estimates for the unknown parameters that are needed for validating the model. While conducting this statistical analysis, we make some key assumptions. Validation comprises of a three-stage statistical procedure. The first stage tests for the multivariate normality,the second stage tests the structure of the covariance matrix between responses, and the third stage tests for the adequacy of the proposed model. The statistical analysis and validation procedures are illustrated with an example of a hospital simulation study. / Master of Science

In a typical response surface study, an experimenter will fit a first order model in the early stages of the study and obtain the path of steepest ascent. The path leads the experimenter out of this initial region of interest and into a new region of interest. The experimenter may fit another first order model here or, if curvature is believed to be present in the underlying system, a second order model. In the final stages of the study, the experimenter fits a second order model and typically contracts the region of interest as the levels of the factors that optimize the response are nearly determined. Due to the sequential nature of experimentation in a typical response surface study, the experimenter may find himself/herself wanting to augment some initial design with additional runs within the current region of interest. The little discussion that exists in the statistical literature suggests adding runs sequentially in a conditional D-optimal manner. Four prediction oriented criteria, I_IV, I_SV_r, I_SV_r^ADJ and G, and two estimation oriented criteria, A and E, are studied here as other possible sequential design augmentation optimality criteria. Analytical properties of I_IV, I_SV_r, and A are developed within the context of the design augmentation problem. I_SV_r is found to be somewhat ineffective in actual sequential design augmentation situations. A new more effective criterion,I_SV_r^ADJ is introduced and thoroughly developed. Software is developed which allows sequential design augmentation via these seven criteria. Unlike existing design augmentation software, all locations within the current region of interest are eligible for inclusion in the augmenting design (a continuous candidate list). Case studies were performed. For a first order model there was negligible difference in the prediction variance properties of the designs generated via sequential augmentation by D and the A best of the other criteria, I_IV, I_SV_r^ADJ, and A. For a second order model, however, the designs generated via sequential augmentation by D place too few runs too late in the interior of the region of interest. Thus, designs generated via sequential augmentation by D yield inferior prediction variance properties to the designs generated via I_IV, I_SV_r^ADJ, and A. The D-efficiencies of the designs generated via sequential augmentation by I_IV, I_SV_r^ADJ, and A range from the reasonable to fully D-optimum. Therefore, the I_IV, I_SV_r^ADJ, optimality criteria are recommended for sequential design augmentation when quality of prediction is more important than quality in estimation of coefficients. / Ph. D.

Statistical research can be more difficult to plan than other kinds of projects, since the research must adapt as knowledge is gained. This dissertation establishes a formal language and methodology for designing experimental research strategies with limited resources. It is a mathematically rigorous extension of a sequential and adaptive form of statistical research called response surface methodology. It uses sponsor-given information, conditions, and resource constraints to decompose an overall project into individual stages. At each stage, a "parent" decision-maker determines what design of experimentation to do for its stage of research, and adapts to the feedback from that research's potential "children", each of whom deal with a different possible state of knowledge resulting from the experimentation of the "parent". The research of this dissertation extends the real-world rigor of the statistical field of design of experiments to develop an deterministic, adaptive algorithm that produces deterministically generated, reproducible, testable, defendable, adaptive, resource-constrained multi-stage experimental schedules without having to spend physical resource.

No description available.

Indiana University-Purdue University Indianapolis (IUPUI) / Long term no-till (NT) farming can improve the CH4 oxidation capacity of agricultural lands through creation of a favorable soil environment for methanotrophs and diffusive gas transport. However, limited data is available to evaluate the merit of that contention. Although the potential for biological CH4 oxidation may exist in NT soils, restricted diffusion could limit expression of that potential in fine-textured soils. A study was conducted to assess the CH4 oxidation potential and gaseous diffusivity of soils under plow till (PT) and NT for > 50 years. Intact cores and composite soils samples (0-10 and 10-20 cm) were collected from NT and PT plots located at a well-drained site (Wooster silt loam) and at a poorly-drained (Crosby silt loam) site in Ohio. Adjacent deciduous forest soils were also sampled to determine maximum rate expected in undisturbed soils in the region. Regardless of study sites and soil depth, CH4 oxidation rate (measured at near ambient CH4) and oxidation potential (Vmax, measured at elevated CH4) were 3-4 and 1.5 times higher in NT than in PT soils, respectively. Activity in the NT soils approached (66-80 %) that in the forest soils. Half saturation constants (Km) and threshold for CH4 oxidation (Th) were lower in NT (Km: 100.5 µL CH4 L-1; Th: 0.5 µL CH4 L-1) than in PT soils (Km: 134 µL CH4 L-1; Th: 2.8 µL CH4 L-1) suggesting a greater affinity of long-term NT soils for CH4, and a possible shift in methanotrophic community composition. CH4 oxidation rates were lower in intact soil cores compared to sieved soils, suggesting that CH4 oxidation was limited by diffusion, a factor that could lead to lower field-measured CH4 uptake than suggested by biological oxidation capacity measured in the laboratory. Regardless of soil drainage characteristic, long-term NT resulted in significantly higher (2-3 times) CH4 diffusivity (mean: 2.5 x 10-3 cm2 s-1) than PT (1.5 x 10-3 cm2 s-1), probably due to improved soil aggregation and greater macro-pores volume in NT soils. Overall, these results confirm the positive impact of NT on the restoration of the biological (Vmax, Km and Th) and physical (diffusivity) soil attributes essential for CH4 uptake in croplands. Long-term implementation of NT farming can therefore contribute to the mitigation of CH4 emission from agriculture.

Indiana University-Purdue University Indianapolis (IUPUI) / Robustness and generalizability of supervised learning algorithms depend on the quality of the labeled data set in representing the real-life problem. In many real-world domains, however, we may not have full knowledge of the underlying data-generating mechanism, which may even have an evolving nature introducing new classes continually. This constitutes a partially-observed setting, where it would be impractical to obtain a labeled data set exhaustively defined by a fixed set of classes. Traditional supervised learning algorithms, assuming an exhaustive training library, would misclassify a future sample of an unobserved class with probability one, leading to an ill-defined classification problem. Our goal is to address situations where such assumption is violated by a non-exhaustive training library, which is a very realistic yet an overlooked issue in supervised learning. In this dissertation we pursue a new direction for supervised learning by defining self-adjusting models to relax the fixed model assumption imposed on classes and their distributions. We let the model adapt itself to the prospective data by dynamically adding new classes/components as data demand, which in turn gradually make the model more representative of the entire population. In this framework, we first employ suitably chosen nonparametric priors to model class distributions for observed as well as unobserved classes and then, utilize new inference methods to classify samples from observed classes and discover/model novel classes for those from unobserved classes. This thesis presents the initiating steps of an ongoing effort to address one of the most overlooked bottlenecks in supervised learning and indicates the potential for taking new perspectives in some of the most heavily studied areas of machine learning: novelty detection, online class discovery and semi-supervised learning.

Indiana University-Purdue University Indianapolis (IUPUI) / Background: Online courses in healthcare programs like Dietetics have increased in availability and popularity. Objective: The purpose of this study was to investigate the connections between online learning environments and Myers-Briggs Type Indicator (MBTI) dimensions among Nutrition Professionals. This research will add to the knowledge base of educators responsible for the design and development of online nutrition courses and will enhance Nutrition Professionals’ academic and professional outcomes. Design: Semi-experimental study design. Subjects/Setting: Thirty-one Nutrition Professionals with mean age of 29 years old. All elements of the study were done online. Statistical Analysis: MBTI dimension summaries were done for descriptive statistics. Fisher’s Exact Test was used to compare frequency of MBTI dimensions in the learning modules (LM) and to analyze learning modality preference based on MBTI dimensions. Two-Sample T-Tests compared test scores for LM groups and test scores for extraverts and introverts. Paired T-Test assessed improvement in test scores related to LM preference. Chi-Square Test compared preferences for the second learning module for both LM groups. Results: The majority of participants’ MBTIs were ESFJ at 35% or ISFJ at 19%. There were more extraverts (71%) compared to introverts (29%). Both LM groups had similar MBTI dimensions. Extraverts and introverts had similar improvements in scores and LM preferences. LM groups performed similarly and in general participants preferred the second learning module they were assigned. Preference for the second LM could be because participants enjoyed the first LM and wanted to learn more information. Both LM groups significantly improved their scores (P=<.0001) in their first and second learning modules regardless of learning module design. Participants were highly motivated to learn as evidenced by their enrollment in this study and completion of 10 hours of learning modules. Motivation to learn may have been the strongest reason performance significantly improved. Conclusion: LM groups significantly improved their LM scores and learned similar amounts. MBTI dimensions extravert and introvert and preferred learning modality had limited impact on performance for this sample of Nutrition Professionals. These results indicate that motivation may be the key to increasing performance in online nutrition courses.