Search Results (1 - 14 of 14 Results)

Sort By  
Sort Dir
 
Results per page  

Han, QiangOn Resilient System Testing and Performance Binning
PhD, University of Cincinnati, 2015, Engineering and Applied Science: Computer Science and Engineering
By allowing timing errors to occur and recovering them on-line, resilient systems are designed to eliminate the frequency or voltage margin to improve circuit performance or reduce power consumption. With the existence of error detection and correction circuits, resilient systems bring about new timing constraints for path delay testing. With the characteristics of allowing timing errors to occur and recovering them on-line, the metrics of resilient system performance are different from traditional circuits, which results in new challenges on resilient system performance binning. Due to these new characteristics of resilient systems, it is essential to develop new testing and binning methodologies for them. In this research, we focus on resilient system testing and performance binning, and attempt to push forward the pace of resilient system commercialization. We make the following contributions. First, we propose a new DFT (design-for-testability) technique, which is able to deal with all different types of timing faults existing in resilient systems, and we develop an efficient test method based on binary search for error collection circuits. Then, a performance binning method based on structural at-speed delay testing is developed for resilient systems to greatly save the binning cost, and an adaptive clock configuration technique is proposed for yield improvement. Last but not least, we propose a new statistical performance analysis tool for resilient systems, called SERA (statistical error rate analysis), which takes process variations into consideration for error rate analysis and produces performance distribution function. With the help of SERA, we develop a profit-oriented binning methodology for resilient systems.

Committee:

Wen-Ben Jone, Ph.D. (Committee Chair); Chien-In Henry Chen, Ph.D. (Committee Member); Harold Carter, Ph.D. (Committee Member); Carla Purdy, Ph.D. (Committee Member); Ranganadha Vemuri, Ph.D. (Committee Member)

Subjects:

Computer Engineering

Keywords:

Resilient computing;Delay testing;Performance binning;Yield improvement;Error rate modeling;Statistical analysis

Onur, Emine MercanPREDICTING THE PERMEABILITY OF SANDY SOILS FROM GRAIN SIZE DISTRIBUTIONS
MS, Kent State University, 2014, College of Arts and Sciences / Department of Geology
Permeability is one of the most important and frequently used properties of soils. Grain size distribution and density are known to influence the permeability of sandy soils. Although the relationships between grain size distribution and permeability has been quantified in previous studies, the influenced of density has not been quantified. The objective of this research was to investigate the quantitative relationships between permeability and grain size distribution indices such as effective particle size (D10), coefficient of uniformity (Cu), coefficient of curvature (Cc), percentage of coarse sand fraction by weight of sample (%C), percentage of medium sand fraction by weight of sample (%M), and percentage of fine sand fraction by weight of sample (%F) to determine whether these relationships could be used for reliable estimates of permeability. Six samples of sandy soils, ranging from well graded to poorly graded, were tested in the laboratory to determine their grain size distribution, maximum dry density (MDD), and optimum water content (OWC). The D10, Cu, Cc, %C, %M, and %F values for each soil were calculated from the grain size distribution plots. Based on the compaction curves, five replicate samples of each soil were prepared at varying dry density values and tested for permeability using the constant head permeability test. Results show that the lowest permeability for sandy soils is achieved at or slightly on the dry side of OWC. To investigate the relationship between permeability and grain size distribution indices, bivariate and step-wise regression analyses were performed. The results show that D10, density, and %M have the strongest correlation (Adjusted R2 = 0.67) with permeability, explaining 67% of the variability in permeability. Permeability depends on the sizes and shapes of interconnections between adjacent pores which, in turn, are influenced by the entire grain size distribution. This research proposes a new grain size distribution index for predicting permeability, designated as the new permeability index. In addition to considering the entire grain size distribution, the new permeability index assigns different weights to different size fractions in the soil with the finest fraction having the maximum weight and the coarsest fraction having the least weight. The new permeability index values for the six soils were correlated with their corresponding permeability values, resulting in a second order quadratic equation with an R2 value of 0.76. This relationship can reliably be used to predict permeability as is indicated by the small amount of residuals between measured and predicted values of permeability. A 3-D model was developed to show the combined effect of the new permeability index and density on permeability.

Committee:

Abdul Shakoor (Advisor)

Subjects:

Geology

Keywords:

Permeability, Grain size distribution, Grain size distribution indices, Density, Sandy soils, Statistical analysis, Permeability index

Fisher, James E.Use of Remote Sensing in the Collection of Discontinuity Data for the Analysis and Design of Cut Slopes
MS, Kent State University, 2011, College of Arts and Sciences / Department of Geology
This study was conducted to examine the use of remote sensing techniques in the collection of discontinuity data for statistical and slope stability analyses. Two study areas where selected in Pulaski and Montgomery counties in central Virginia. Terrestrial LiDAR (light detection and ranging) and a transit compass were used to collect data at an abandoned quarry in the vicinity of Claytor Dam and Interstate 81 southwest of Christiansburg, Virginia. These data were used in a statistical analysis to compare both datasets and in a slope stability analysis for the adjacent section of Interstate 81. Digital photogrammetry was used to collect data on slopes along Interstate 81 northeast of Christiansburg. The digital photogrammetry dataset was qualitatively compared with the LiDAR dataset to illustrate differences and possible limitations of these remote sensing methods for the collection of discontinuity data. The objectives of this study were as follows: 1) compare the use of LiDAR and transit compass methods in collecting discontinuity orientation data through graphical and statistical analyses; 2) compare the kinematic analyses for both LiDAR and transit compass methods to determine the differences in the results; 3) compare LiDAR and photogrammetry methods to evaluate any limitations therein; and 4) compare the use of LiDAR and transit compass methods in the design of cut slopes along a portion of Interstate 81. For the comparison of the LiDAR and transit compass datasets, results show that the two datasets have similar mean orientation values for the corresponding discontinuity sets and are graphically similar when plotted on stereonet plots. However, the two datasets are not statistically derived from the same population. More importantly, a joint set was identified in the transit compass dataset that was either not detected or has a different mean orientation in the LiDAR dataset. These differences affected the kinematic analysis results and, therefore, the cut slope design for Interstate 81. A possible explanation for these differences is that the tectonically disturbed nature of the bedrock within the site area coupled with method limitations for both the LiDAR and the transit compass resulted in the sampling of different subpopulations within the global population. Specifically, for the apparent missing discontinuity set in the LiDAR data, limitations in the method resolution lead to the under-sampling of discontinuity data comprising that discontinuity set, causing it to be poorly represented in the stereonet plots. The qualitative comparison of LiDAR and photogrammetry methods demonstrates that the stereonet plots have distinct differences in terms of discontinuity set data density and general scatter of the data. These differences are due to the distinct data acquisition procedures and processing steps of each method, which illustrates the limitations each method has with respect to collecting and deriving discontinuity orientation data. However, with adequate preparation and planning of field investigations to minimize the effect of method limitations, incorporating these remote sensing techniques will enable faster, more efficient, and safer data collection of discontinuity data for the design of cut slopes.

Committee:

Abdul Shakoor, PhD (Advisor); Donna Witter, PhD (Committee Member); Dahl Peter, PhD (Committee Member)

Keywords:

remote sensing; terrestrial LiDAR; photogrammetry; disconinuity; stereonet; statistical analysis; slope stability; kinematic analysis; engineering geology; Virginia

Madaris, Aaron T.Characterization of Peripheral Lung Lesions by Statistical Image Processing of Endobronchial Ultrasound Images
Master of Science in Biomedical Engineering (MSBME), Wright State University, 2016, Biomedical Engineering
This thesis introduces the concept of implementing greyscale analysis, also known as intensity analysis, on endobronchial ultrasound (EBUS) images for the purposes of diagnosing peripheral lung tumors. The statistical methodology of using greyscale and histogram analysis allows the characterization of lung tissue in EBUS images. Regions of interest (ROI) will be analyzed in MATLAB and a feature vector will be created. A feature vector of first-order, second-order and histogram greyscale analysis will be created and used for the classification of malignant vs benign peripheral lung tumors. The tools that were implemented were MedCalc for the initial statistical analysis of receiver operating curves (ROC), Multiple Regression and MATLAB for the machine learning and ROI collection. Feature analysis, multiple regression and machine learning methods were used to better classify the malignant and benign EBUS images. The classification is assessed with a confusion matrix, ROC curve, accuracy, sensitivity and specificity. It was found that minimum pixel value, contrast and energy are the best determining factors to discriminate between benign and malignant EBUS images.

Committee:

Ulas Sunar, Ph.D. (Advisor); Jason Parker, Ph.D. (Committee Member); Jaime Ramirez-Vick, Ph.D. (Committee Member)

Subjects:

Biomedical Engineering; Biomedical Research; Biostatistics; Computer Engineering; Engineering; Health Care; Medical Imaging

Keywords:

Endobronchial Ultrasound; Medical Imaging; Image Analysis; Statistical Analysis; Machine learning; MATLAB; Histogram; Texture; Multiple Regression; Feature analysis

Fourman, Jeffrey F.When Insurgents Go Terrorist: The Role of Foreign Support in the Adoption of Terrorism
Master of Arts (MA), Wright State University, 2014, International and Comparative Politics
What role does foreign support play when an insurgent group adopts terrorism? Utilizing both quantitative analysis and in-depth case studies, this thesis examines the effects of foreign support among other commonly cited explanations for an insurgency's adoption of terrorism. In addition to observing the effects of foreign support on the adoption of terrorism, the effects of government regime type, insurgent group goal type, insurgent group strength, and foreign benefactor type are analyzed. After executing a multiple logistic regression analysis of 109 intrastate conflicts occurring from 1972 to 2007 and conducting detailed case studies for the Tamils in Sri Lanka and the Kurds in Iraq, this thesis concludes that specific types of foreign support from non-state actors not only make insurgent groups significantly stronger but also make them more likely to adopt terrorism thus calling into question the weapon of the weak argument.

Committee:

Vaughn Shannon, Ph.D. (Committee Chair); Pramod Kantha, Ph.D. (Committee Member); R. William Ayres, Ph.D. (Committee Member)

Subjects:

Comparative; International Relations; Political Science

Keywords:

Insurgency; Terrorism; Foreign Support; Political Violence; Sri Lanka; Tamil; Iraq; Kurd; Statistical Analysis; Indian Peace Keeping Forces; Kurdistan Democratic Party; Patriotic Union of Kurdistan; Liberation Tigers of Tamil Eelam

Colorado Urrea, Gabriel J.Service Life of Concrete and Metal Culverts Located in Ohio Department of Transportation Districts 9 and 10
Master of Science (MS), Ohio University, 2014, Civil Engineering (Engineering and Technology)
In this study, in-service conditions were evaluated to estimate the service life of concrete and metal culverts. The Ohio Research Institute for Transportation and the Environment (ORITE) and a private consulting company proposed new inspection methods and rating procedures for concrete, metal and thermoplastic pipes; concrete and metal culverts are addressed in this study. The inspection activities were developed in culverts located in Ohio Department of Transportation (ODOT) Districts 9 and 10 since the aggressive environmental conditions found in these portions of the state of Ohio. Before each field trip, culverts were selected to meet requirements of location, material, and dimensions. For dimensions, screening criteria were than 42 inch span and rise dimensions, and a maximum length of 150 feet. From each culvert, basic information was gathered from the inventory data provided by ODOT and in the field. The data gathered from the inventory and the field work was statistically analyzed to identify significant factors that contribute to material deterioration. The rating scales proposed by the ORITE and ODOT were employed in the statistical regressions as outcome variables, to measure the effectiveness and accuracy in predicting the remaining service life. Multivariable linear and nonlinear regression models were proposed to estimate the remaining service life of existing metal and concrete structures with similar conditions in the state of Ohio. Results for concrete culverts show that the multivariable linear regression results showed that pH and resistivity of water were significant for the ODOT and ORITE rating scale but the linear model is not since the age is not included in the regression. While, the multivariable non-linear regression results indicated that pH of water, age and span were significant based on the ODOT rating scale. For metal culverts, the multivariable linear regression results showed that rise, span, age, level of abrasion, thickness of the plate, slope, velocity, and depth of the flow were significant based on the ODOT rating scale. And, age, soil cover, level of abrasion, pH of water, thickness of the plate, slope, flow velocity, and depth of flow were all significant for the ORITE rating scale, both models are not practical in estimating the deterioration of metal pipes. Non-linear regression did not generate more reliable results in predicting the service life of metal culverts.

Committee:

Shad Sargand, PhD (Advisor); Teruhisa Masada, PhD (Committee Member); McAvoy Deborah, PhD (Committee Member); Mohlenkamp Martin, PhD (Committee Member)

Subjects:

Civil Engineering; Geotechnology

Keywords:

Concrete and metal culverts; service life; field inspection work; statistical analysis

Niti, DuggalRetail Location Analysis: A Case Study of Burger King & McDonald’s in Portage & Summit Counties, Ohio
MA, Kent State University, 2007, College of Arts and Sciences / Department of Geography
There has been a growing interest among the academia and the private sector for the use of GIS techniques in the analysis and planning of retail store network. Over the past few decades the methodologies used for research of sighting of retail outlets have become more sophisticated as a result of applicable modeling procedures being developed with GIS. This study conducts a retail location analysis of the relationship between the fast-food store performance of McDonald’s and Burger King and the various spatial and socio-economic factors of their respective catchment areas. Analytical procedures in GIS and statistical techniques have been applied to carry out the analysis in this study. Study areas have been partitioned into a set of Thiessen polygons and into various spatial configurations using variable buffer polygons to emulate various spatial configurations of catchment areas (i.e., trade areas) associated with each fast food store. The socio-economic profiles in the partitioned polygons have been analyzed with a series of regression models. The result of the study has brought out a better understanding of how location factors influence the performance of the stores as well as how the socio-economic attributes of the catchment areas affect the store revenues.

Committee:

Jay Lee (Advisor)

Keywords:

Retail Location Analysis; geographic information systems (GIS); Statistical Analysis; Regression Analysis; Geocoding; Catchment Area Analysis; Buffer Polygons; Thiessen Polygons; McDonald's; Burger King; Fast Food Restaurants; Overlay Analysis

Yazawa, DaigoEnvironment Change: An Analysis of College Football Operations
MA, Kent State University, 2014, College and Graduate School of Education, Health and Human Services / School of Foundations, Leadership and Administration
This research examines environment changes that occur in contemporary college athletics. Factors that impact internal and external operations, specific to college football programs, will be explored. For this study, environment changes will include assessment specific to facility upgrades/renovations, new coaches, and conference moves. This research utilizes Point Biserial Correlation Tests and Paired-Samples T-Tests to investigate relationships between those three variables and recruiting/home game attendance.

Committee:

Mark Lyberger, Ph.D. (Committee Chair); Aaron Mulrooney, J.D. (Committee Member); Jian Li, Ph.D. (Committee Member)

Subjects:

Sports Management; Statistics

Keywords:

Environment Changes; College Football Operations; Statistical Analysis; College Football Arms Race; Conference Realignment; Facility Upgrades and Renovations; Recruiting; Home Game Attendance

Mahadevan Muralidharan, AnanthAnalysis of Garbage Collector Algorithms in Non-Volatile Memory Devices
Master of Science, The Ohio State University, 2013, Computer Science and Engineering
Non-volatile memory devices or flash, even with many advantages, still have a few problems such as the inability to update data in place. This necessitates the need for a garbage collector (GC) that can collect active data and create space by erasing flash blocks. However this is a very costly operation that increases the write latency thereby lowering the efficiency of the flash device. The frequency at which the GC is invoked by the underlying file system depends on the data’s traffic pattern as well as the fullness of the device. It is therefore important to study different GC algorithms for different traffic patterns and at varying fullness levels in order to find the most efficient one for a particular situation. In this report we study the efficiency of byte address non-volatile memory devices (such as NOR), under varying traffic patterns. We study the algorithms using simulations coded in Matlab. A simulator for the flash file system as well as the GC algorithms and various applications traffic was developed and used for the study. We compare and contrast the efficiency and the time taken for the GCs at utilization levels ranging from 2% to 98%. We also model some of the algorithms analytically and find that our analytical results match our simulations. The performance results for five different GC algorithms for flash devices for three traffic/access patterns are presented in this report. The access patterns include long-tailed, uniform and bimodal distributions. The algorithms studied are a round-robin style first in first out (FIFE), a greedy least active clean (LAC), 3-Generation (3-Gen) GC, N-Generation (N-Gen) GC (a generalized generation algorithm) and Eta-N-Generation (Eta-N-Gen) GC (a variation on N-Gen). The results indicate that round-robin style GC algorithm (FIFE) and greedy algorithm (LAC) perform better in most of the scenarios than generational algorithms. This is counter-intuitive to the existing norms. LAC slightly underperforms the FIFE under heavy flash utilization. For long-tailed traffic – the canonical use case for generational algorithms – FIFE and LAC still perform better than generational algorithms. The reason is that, it is non-trivial to configure a generational algorithm to get the optimum performance for a particular traffic pattern. To optimize performance, the radio of the size of subsequent generations should be the same as ratio between cold data and the rest of the data. Since in most application cases we do not know this a priori, static optimal configuration of generational algorithms is impossible. However an adaptive algorithm which changes allocations between generations on the fly could achieve better efficiency. Further we find that for better efficiency, at low levels of utilization it is important to isolate “cold’ data well, but at higher utilization identifying and handling hot data (i.e., never move the hot data) is important. Results from our study suggest that FIFE might work well for most of the application scenarios.

Committee:

Rajiv Ramnath (Advisor); Jayashree Ramanathan (Committee Member)

Subjects:

Computer Engineering; Computer Science

Keywords:

Garbage collection; Non-volatile memory device; Solid state device; statistical analysis; Uniform distribution; Pareto distribution; Bimodal distribution; Generational garbage collection; Greedy garbage collection; efficient garbage collection;

McCardy, Nicole RPrediction of Surfactant Mildness for Rinse-off Formulations Using Preclinical Assays
MS, University of Cincinnati, 2016, Pharmacy: Pharmaceutical Sciences
Mixed surfactant and surfactant–polymer compositions have been reported to decrease surfactant deposition onto and into the skin relative to single surfactant compositions, potentially improving the mildness of the product. Previous workers in this area (see Moore et al., J Cosmet Sci 54:29–46 (2003) and subsequent publications), employed a procedure in which excised porcine skin was exposed to a surfactant solution containing radiolabeled sodium dodecyl sulfate ((14)C–SDS) for 5 hours. We have developed an improved (14)C–SDS deposition assay using excised human skin that reflects typical consumer exposure times for rinse–off products. Using the new protocol, we were able to see a significant decrease in (14)C–SDS deposition from an SDS⁄PEG 8000 composition applied to excised skin for either 2 minutes or 10 minutes, as compared with SDS only. Following this, a study was designed to determine whether data from a carefully designed array of preclinical assays could effectively predict the harshness of mixed surfactant and surfactant–polymer compositions on human skin, as measured by corneometry and visual dryness scores in a five–day forearm controlled application test (FCAT). The test compositions included surfactants commonly used in rinse–off applications including shampoos and shower gels. A total of seventeen compositions were tested. The preclinical methods included the recently–developed surfactant deposition assay, zein solubilization, stearic acid solubilization, micelle size, and critical micelle concentration (CMC). The changes–from–baseline (CFB) of the two primary clinical measures, corneometer reading and expert–assessed visual dryness score, were analyzed in terms of the preclinical assay results according to linear regression for bivariate analyses and partial least squares (PLS) for multivariate analyses. Cross–validation was performed within PLS via a leave–one–out algorithm in order to prevent overfitting of the clinical data. FCAT test results correlated significantly with surfactant deposition (corneometer: r2 = 0.631, visual dryness: r2 = 0.498), micelle size (corneometer: r 2 = 0.551, visual dryness: r 2 = 0.445) and zein solubilization (corneometer: r 2 = 0.480, visual dryness: r 2 = 0.145). A one–component PLS model using normalized and scaled data from three of the five preclinical assays — surfactant deposition, micelle size and zein solubilization — yielded the strongest correlations (corneometer: r 2 = 0.889, visual dryness: r 2 = 0.861). Milder formulations were associated with lower surfactant deposition, larger micelle size, and lower zein solubilization. The study results show that, within the composition range tested, preclinical assay data can be strongly correlated to clinical measures of skin dryness. The results support the hypothesis that micellar structure is more important to surfactant mildness than is CMC, with larger micelles leading to milder formulations.

Committee:

Gerald Kasting, Ph.D. (Committee Chair); Harshita Kumari, Ph.D. (Committee Member); Ryan Thompson, B.S. (Committee Member)

Subjects:

Pharmaceuticals

Keywords:

colloid science;surface chemistry;stratum corneum;skin science;surfactant mildness;multivariate statistical analysis

Engle, KevinA LATE GLACIAL-EARLY HOLOCENE PALEOCLIMATE SIGNAL FROM THE OSTRACODE RECORD OF TWIN PONDS, VERMONT
MS, Kent State University, 2015, College of Arts and Sciences / Department of Geology
During the fall of 2012, a sediment core was collected from Twin Ponds Lake in Brookfield, Vermont. Twin Ponds Lake is located in central Vermont with the surrounding rocks being Silurian to Devonian age phyllite and limestone. This research is focusing on using isotopic analysis of ostracodes from the core in order to assist in climatic reconstruction of the region from the Late Glacial to the Early Holocene. The ostracodes are part of a multi-university effort to analyze the core: additional research is focusing on pollen, loss on ignition, charcoal and bulk carbonate analysis of the core. New England is very sensitive to climate change so isotopic analysis of the carbonate in the ostracode carapaces along with the bulk carbonate analysis will provide a very detailed account of the changing climate of the region. The sediment record of this core ranges in time from the Late Glacial throughout the Holocene but the most detailed records are found from just before and after the Younger Dryas. This study is important for the ostracode record, being the first Holocene ostracode record for New England outside of Lake Champlain. This record shows the easternmost distribution of the species from the mid-continent such as Candona ohioensis, Candona paraohioensis and Candona candida. Cyclocypris globosa is a cold tolerant species, most commonly found in the Yukon. Its presence towards the bottom of the core is a clear Younger Dryas signal. Additionally, the presence and abundance of certain species of ostracodes are good indicators of a rising and falling lake level.

Committee:

Alison Smith (Advisor); Donald Palmer (Committee Member); Joseph Ortiz (Committee Member)

Subjects:

Geochemistry; Geology; Paleoclimate Science; Paleoecology

Keywords:

Ostracodes, Paleoclimate, Micropaleontology, Paleolimnology, Geology, Oxygen Isotopes, Statistical Analysis

Panozzo, Kimberly AA Validation of Nass Crop Data Layer in the Maumee River Watershed
Master of Arts, University of Toledo, 2016, Geography
It is suspected that corn and soybean production in the Maumee Watershed has contributed to nutrient loading into Lake Erie, therefore affecting the frequency and duration of toxic algae (Dolan 1993), (Michalak, et al. 2012). Accurate crop type estimation is important in order to determine the potential impact on the lake and assess methods to reduce excess nutrient loading. Increasingly, the National Agricultural Statistics Survey (NASS) Crop Data Layer (CDL) is being used as a primary input for agricultural research and crop estimation modeling therefore assessing the accuracy of the CDL is imperative. This study aims to validate the CDL, assess accuracy differences on multiple spatial scales and to examine the efficiencies of using the CDL for future research in the region. Results of CDL validation using in situ field observations from 2011 and 2012 indicate an overall accuracy of at 94% and 92% respectively and khat accuracy of 90% (2011) and 86% (2012). Crop specific accuracy for corn, soy and wheat also resulted in considerably high user accuracy values, with slight differences between years. Accuracy measures vary by region and by year however in each circumstance analyzed, the differences were not significant. Of these measureable difference, it was shown that the 2012 comparison contained a higher degree of difference and this may be attributed to drought in the region for this year. It is concluded that NASS’s CDL is an effective and efficient product for agricultural research.

Committee:

Kevin Czajkowski, PHD (Committee Chair); P.L. Lawrence, PHD (Committee Member); Dan Hammel, PHD (Committee Member)

Subjects:

Agriculture; Geographic Information Science; Geography; Land Use Planning; Remote Sensing

Keywords:

Crop Data Layer; NASS; Remote Sensing; Agriculture; Agricultural Land Use; Thematic map validation; Validation; Accuracy Assessment; Statistical Analysis; Comparative Analysis; Maumee River; Maumee River Watershed

Wang, TaoStatistical design and analysis of microarray experiments
Doctor of Philosophy, The Ohio State University, 2005, Statistics
Microarray, a bio-technology that allows monitoring of gene expressions for thousands of genes simultaneously, has revolutionized biological and genomic research and holds promising potentials in many real applications, such as drug targeting, gene profiling, disease diagnosis and prognosis, pharmacogenomics, etc. Along with its unprecedented potential, microarray technology presents miscellaneous challenges in statistical analysis of microarray gene expression data. Many sources of extraneous variations are present in a microarray experiment. Adjusting these extraneous variations is critical to the separation of biological signals from artifacts. Moreover, microarray gene expression data typically are of extremely large dimension, consisting of tens of thousands of observations. Computational efficiency in statistical analysis is therefore crucial. For testing the significance of biological signal, multiplicity adjustment is indispensable. We propose a modeling approach that allows flexible experimental design, while providing accurate estimation and easy multiplicity-adjusted inferences. This modeling approach is suitable for various types of microarrays, including both cDNA and oligonucleotide microarrays. The statistical modeling and multiplicity-adjusted inference are integrated into an R package, MultiArray , as a computationally efficient environment. Real microarray experiment examples show that our modeling approach and MultiArray outperform other popular packages in both detecting differences and establishing equivalence in gene expressions.

Committee:

Jason Hsu (Advisor)

Subjects:

Statistics

Keywords:

microarray; gene expression; statistical analysis; modeling; statistical design

Goodpaster, Aaron M.Statistical Analysis Methods Development for Nuclear Magnetic Resonance and Liquid Chromatography/Mass Spectroscopy Based Metabonomics Research
Doctor of Philosophy, Miami University, 2011, Chemistry
This dissertation describes new statistical analysis methods for nuclear magnetic resonance (NMR) spectroscopy and liquid chromatography/mass spectroscopy (LC/MS) – based metabonomics along with determining if diapers and cotton balls could introduce contamination into newborn babies urine when collecting their urine for metabonomic studies using the “cotton ball in diaper” method of collection. Chapter 1 provides a background of metabonomics along with the advantages and disadvantages of the two analytical techniques used, NMR and LC/MS, along with how the data is analyzed using multivariate data analysis. Chapter 2 develops a decision tree algorithm to determine which buckets in the loadings plot are statistically significant in the principal component analysis (PCA). After the statistical analysis the loadings plot was colored based on a “heat-map” according to the p-score for each loading, which showed how significant the buckets were in the analysis. Based on the significance analysis, the effect size was calculated to allow for the calculation of the study size that is needed for each group in a metabonomics study. Chapter 3 develops a metric for quantification of cluster separation and for assessment of the statistical significance of cluster separation seen in the scores plot of the PCA. Along with this it was shown how Pareto scaling and partial least squares – discriminant analysis (PLS-DA) affected the separation in the scores plot, along with how it affects the loadings plot of a PCA. Chapter 4 investigated the coefficient of variation (CV) of NMR data collected over an eight-month period and how the signal-to-noise (S/N) ratio affected the CV. This study determined that the CV has no relation to the S/N ratio and that the CV must be considered when validating a potential biomarker. Chapter 5 investigates the potential effect of diaper and cotton ball contamination on NMR and LC/MS-based metabonomic studies of urine from newborn babies. It was determined that the diaper could potentially affect a metabonomic study, but when the cotton ball is inserted the chance for contamination decreases. Chapter 6 summarizes the presented research.

Committee:

Michael A. Kennedy, PhD (Advisor); Neil D. Danielson, PhD (Committee Chair); Carole Dabney-Smith, PhD (Committee Member); Blanton S. Tolbert, PhD (Committee Member); Natosha L. Finley, PhD (Committee Member)

Subjects:

Analytical Chemistry; Biostatistics; Chemistry

Keywords:

Metabonomics; Statistical Analysis; Nuclear Magnetic Resonance; Liquid Chromatography; Mass Spectroscopy; NMR; LC/MS