Digital Commons Network^™

Quantifying The Effects Of Correlated Covariates On Variable Importance Estimates From Random Forests, Ryan Vincent Kimes

Theses and Dissertations

Recent advances in computing technology have lead to the development of algorithmic modeling techniques. These methods can be used to analyze data which are difficult to analyze using traditional statistical models. This study examined the effectiveness of variable importance estimates from the random forest algorithm in identifying the true predictor among a large number of candidate predictors. A simulation study was conducted using twenty different levels of association among the independent variables and seven different levels of association between the true predictor and the response. We conclude that the random forest method is an effective classification tool when the goals …

Design And Analysis Methods For Cluster Randomized Trials With Pair-Matching On Baseline Outcome: Reduction Of Treatment Effect Variance, Misook Park

Theses and Dissertations

Cluster randomized trials (CRT) are comparative studies designed to evaluate interventions where the unit of analysis and randomization is the cluster but the unit of observation is individuals within clusters. Typically such designs involve a limited number of clusters and thus the variation between clusters is left uncontrolled. Experimental designs and analysis strategies that minimize this variance are required. In this work we focus on the CRT with pre-post intervention measures. By incorporating the baseline measure into the analysis, we can effectively reduce the variance of the treatment effect. Well known methods such as adjustment for baseline as a covariate …

Assessing, Modifying, And Combining Data Fields From The Virginia Office Of The Chief Medical Examiner (Ocme) Dataset And The Virginia Department Of Forensic Science (Dfs) Datasets In Order To Compare Concentrations Of Selected Drugs, Amy Elizabeth Herrin

Theses and Dissertations

The Medical Examiner of Virginia (ME) dataset and the Virginia Department of Forensic Science Driving Under the Influence of Drugs (DUI) datasets were used to determine whether people have the potential to develop tolerances to diphenhydramine, cocaine, oxycodone, hydrocodone, methadone, and morphine. These datasets included the years 2000-2004 and were used to compare the concentrations of these six drugs between people who died from a drug-related cause of death (of the drug of interest) and people who were pulled over for driving under the influence. Three drug pattern groups were created to divide each of the six drug-specific datasets in …

A Comparison For Longitudinal Data Missing Due To Truncation, Rong Liu

Theses and Dissertations

Many longitudinal clinical studies suffer from patient dropout. Often the dropout is nonignorable and the missing mechanism needs to be incorporated in the analysis. The methods handling missing data make various assumptions about the missing mechanism, and their utility in practice depends on whether these assumptions apply in a specific application. Ramakrishnan and Wang (2005) proposed a method (MDT) to handle nonignorable missing data, where missing is due to the observations exceeding an unobserved threshold. Assuming that the observations arise from a truncated normal distribution, they suggested an EM algorithm to simplify the estimation.In this dissertation the EM algorithm is …

Algebraic Characterizations Of Graph Imbeddability In Surfaces And Pseudosurfaces, Lowell Abrams, Dan Slilaty

Mathematics and Statistics Faculty Publications

Given a finite connected graph G and specifications for a closed, connected pseudosurface, we characterize when G can be imbedded in a closed, connected pseudosurface with the given specifications. The specifications for the pseudosurface are: the number of face-connected components, the number of pinches, the number of crosscaps and handles, and the dimension of the first Z2-homology group. The characterizations are formulated in terms of the existence of a dual graph G ∗ on the same set of edges as G which satisfies algebraic conditions inspired by homology groups and their intersection products.

On Adaptive Testing In Orthogonal Saturated Designs, Daniel T. Voss, Weizhen Wang

Mathematics and Statistics Faculty Publications

Adaptive, size-a step-down tests are provided for the analysis of orthogonal saturated designs. The tests work effectively under effect sparsity, and include as special cases the individual nonadaptive tests of Berk and Picard (1991) and the simultaneous nonadaptive tests of Voss (1988). The approach is similar to that used by Wang and Voss (2003) to construct adaptive confidence intervals, but testing is simpler because one can use the same denominator for all statistics. Step-down tests also have a clear power advantage over simultaneous confidence intervals and analogous single-step tests, as is demonstrated theoretically and assessed via simulation.

Two-Dimensional Penalized Signal Regression For Hand Written Digit Recognition, Qing Tang

LSU Master's Theses

Many attempts have been made to achieve successful recognition of handwritten digits. We report our results of using statistical method on handwritten digit recognition. A digitized handwritten numeral can be represented by an image with grayscales. The image includes features that are mapped into two-dimensional space with row and column coordinates. Based on this structure, two-dimensional penalized signal logistic regression (PSR) is applied to the recognition of handwritten digits. The data set is taken from the USPS zip code database that contains 7219 training images and 2007 test images. All the images have been deslanted and normalized into 16 x …

Optimal Clustering: Genetic Constrained K-Means And Linear Programming Algorithms, Jianmin Zhao

Theses and Dissertations

Methods for determining clusters of data under- specified constraints have recently gained popularity. Although general constraints may be used, we focus on clustering methods with the constraint of a minimal cluster size. In this dissertation, we propose two constrained k-means algorithms: Linear Programming Algorithm (LPA) and Genetic Constrained K-means Algorithm (GCKA). Linear Programming Algorithm modifies the k-means algorithm into a linear programming problem with constraints requiring that each cluster have m or more subjects. In order to achieve an acceptable clustering solution, we run the algorithm with a large number of random sets of initial seeds, and choose the solution …

Synthesis V. Purity And Large-N Studies: How Might We Assess The Gap Between Promise And Performance?, Will H. Moore

Human Rights & Human Welfare

A review of:

Protecting Human Rights: A Comparative Study by Todd Landman. Washington, DC: Georgetown University Press, 2005, 231 pp.

A Class Of Nonlinear Stochastic Volatility Models, Jun Yu, Zhenlin Yang

Research Collection School Of Economics

This paper proposes a class of nonlinear stochastic volatility models based on the Box-Cox transformation which offers an alternative to the one introduced in Andersen (1994). The proposed class encompasses many parametric stochastic volatility models that have appeared in the literature, including the well known lognormal stochastic volatility model, and has an advantage in the ease with which different specifications on stochastic volatility can be tested. In addition, the functional form of transformation which induces marginal normality of volatility is obtained as a byproduct of this general way of modeling stochastic volatility. The efficient method of moments approach is used …

Methods For The Estimation Of Missing Values In Time Series, David S. Fung

Theses: Doctorates and Masters

Time Series is a sequential set of data measured over time. Examples of time series arise in a variety of areas, ranging from engineering to economics. The analysis of time series data constitutes an important area of statistics. Since, the data are records taken through time, missing observations in time series data are very common. This occurs because an observation may not be made at a particular time owing to faulty equipment, lost records, or a mistake, which cannot be rectified until later. When one or more observations are missing it may be necessary to estimate the model and also …

Semiparametric Approaches For Joint Modeling Of Longitudinal And Survival Data With Time Varying Coefficients, Xiao Song, C.Y. Wang

UW Biostatistics Working Paper Series

We study joint modeling of survival and longitudinal data. There are two regression models of interest. The primary model is for survival outcomes, which are assumed to follow a time varying coefficient proportional hazards model. The second model is for longitudinal data, which are assumed to follow a random effects model. Based on the trajectory of a subject's longitudinal data, some covariates in the survival model are functions of the unobserved random effects. Estimated random effects are generally different from the unobserved random effects and hence this leads to covariate measurement error. To deal with covariate measurement error, we propose …

Nonparametric Control Chart For The Range, Arnold J. Stromberg

Statistics Faculty Patents

A method is provided for detecting or predicting an undesired deviation in variability of at least one parameter being monitored, wherein the variation in the parameter is incrementally recorded. The method comprises establishing the number of subsets of a dataset that have a range of the difference between any two datapoints within the dataset, and computing a control chart for the range based thereon. The method accurately detects changes in variability in real time. The true distribution of the data is reflected, and the desired result is achieved without requiring an inordinate number of computations.

Foreign Migration To The Cleveland-Akron-Lorain Metropolitan Area From 1995 To 2000, Mark Salling, Ellen Cyran

All Maxine Goodman Levin School of Urban Affairs Publications

This report is one of a series on migration to and from the region using the five percent Public Use Microdata Sample (PUMS) of the 2000 Census of Population and Housing and provides a description of foreign migrants moving to the Cleveland-Akron-Lorain (CAL) Consolidated Metropolitan Area (CMSA) from 1995 to 2000.* The report identifies the countries of origin of migrants and compares the demographic, socioeconomic, and housing characteristics of the foreign migrants to the CAL with other groups, including foreign migrants to Ohio and the nation, and, at times, to domestic migrants to and from the CAL.

Alleviating Linear Ecological Bias And Optimal Design With Subsample Data, Adam Glynn, Jon Wakefield, Mark Handcock, Thomas Richardson

UW Biostatistics Working Paper Series

In this paper, we illustrate that combining ecological data with subsample data in situations in which a linear model is appropriate provides three main benefits. First, by including the individual level subsample data, the biases associated with linear ecological inference can be eliminated. Second, by supplementing the subsample data with ecological data, the information about parameters will be increased. Third, we can use readily available ecological data to design optimal subsampling schemes, so as to further increase the information about parameters. We present an application of this methodology to the classic problem of estimating the effect of a college degree …

Bayesian Analysis Of Cell-Cycle Gene Expression Data, Chuan Zhou, Jon Wakefield, Linda Breeden

UW Biostatistics Working Paper Series

The study of the cell-cycle is important in order to aid in our understanding of the basic mechanisms of life, yet progress has been slow due to the complexity of the process and our lack of ability to study it at high resolution. Recent advances in microarray technology have enabled scientists to study the gene expression at the genome-scale with a manageable cost, and there has been an increasing effort to identify cell-cycle regulated genes. In this chapter, we discuss the analysis of cell-cycle gene expression data, focusing on a model-based Bayesian approaches. The majority of the models we describe …

Empirical Likelihood Inference For The Area Under The Roc Curve, Gengsheng Qin, Xiao-Hua Zhou

UW Biostatistics Working Paper Series

For a continuous-scale diagnostic test, the most commonly used summary index of the receiver operating characteristic (ROC) curve is the area under the curve (AUC) that measures the accuracy of the diagnostic test. In this paper we propose an empirical likelihood approach for the inference of AUC. We first define an empirical likelihood ratio for AUC and show that its limiting distribution is a scaled chi-square distribution. We then obtain an empirical likelihood based confidence interval for AUC using the scaled chi-square distribution. This empirical likelihood inference for AUC can be extended to stratified samples and the resulting limiting distribution …

Interval Estimation For The Ratio And Difference Of Two Lognormal Means, Yea-Hung Chen, Xiao-Hua Zhou

UW Biostatistics Working Paper Series

Health research often gives rise to data that follow lognormal distributions. In two sample situations, researchers are likely to be interested in estimating the difference or ratio of the population means. Several methods have been proposed for providing confidence intervals for these parameters. However, it is not clear which techniques are most appropriate, or how their performance might vary. Additionally, methods for the difference of means have not been adequately explored. We discuss in the present article five methods of analysis. These include two methods based on the log-likelihood ratio statistic and a generalized pivotal approach. Additionally, we provide and …

Inferences In Censored Cost Regression Models With Empirical Likelihood, Xiao-Hua Zhou, Gengsheng Qin, Huazhen Lin, Gang Li

UW Biostatistics Working Paper Series

In many studies of health economics, we are interested in the expected total cost over a certain period for a patient with given characteristics. Problems can arise if cost estimation models do not account for distributional aspects of costs. Two such problems are 1) the skewed nature of the data and 2) censored observations. In this paper we propose an empirical likelihood (EL) method for constructing a confidence region for the vector of regression parameters and a confidence interval for the expected total cost of a patient with the given covariates. We show that this new method has good theoretical …

Confidence Intervals For Predictive Values Using Data From A Case Control Study, Nathaniel David Mercaldo, Xiao-Hua Zhou, Kit F. Lau

UW Biostatistics Working Paper Series

The accuracy of a binary-scale diagnostic test can be represented by sensitivity (Se), specificity (Sp) and positive and negative predictive values (PPV and NPV). Although Se and Sp measure the intrinsic accuracy of a diagnostic test that does not depend on the prevalence rate, they do not provide information on the diagnostic accuracy of a particular patient. To obtain this information we need to use PPV and NPV. Since PPV and NPV are functions of both the intrinsic accuracy and the prevalence of the disease, constructing confidence intervals for PPV and NPV for a particular patient in a population with …

Model Checking For Roc Regression Analysis, Tianxi Cai, Yingye Zheng

Harvard University Biostatistics Working Paper Series

The Receiver Operating Characteristic (ROC) curve is a prominent tool for characterizing the accuracy of continuous diagnostic test. To account for factors that might invluence the test accuracy, various ROC regression methods have been proposed. However, as in any regression analysis, when the assumed models do not fit the data well, these methods may render invalid and misleading results. To date practical model checking techniques suitable for validating existing ROC regression models are not yet available. In this paper, we develop cumulative residual based procedures to graphically and numerically assess the goodness-of-fit for some commonly used ROC regression models, and …

On The Use Of Non-Euclidean Isotropy In Geostatistics, Frank C. Curriero

Johns Hopkins University, Dept. of Biostatistics Working Papers

This paper investigates the use of non-Euclidean distances to characterize isotropic spatial dependence for geostatistical related applications. A simple example is provided to demonstrate there are no guarantees that existing covariogram and variogram functions remain valid (i.e.\ positive definite or conditionally negative definite) when used with a non-Euclidean distance measure. Furthermore, satisfying the conditions of a metric is not sufficient to ensure the distance measure can be used with existing functions. Current literature is not clear on these topics. There are certain distance measures that when used with existing covariogram and variogram functions remain valid, an issue that is explored. …

Autologous Stem Cell Transplant: Factors Predicting The Yield Of Cd34+ Cells, Elizabeth Anne Lawson

Theses and Dissertations

Stem cell transplant is often considered the last hope for the survival for many cancer patients. The CD34+ cell content of a collection of stem cells has appeared as the most reliable indicator of the quantity of desired cells in a peripheral blood stem cell harvest and is used as a surrogate measure of the sample quality. Factors predicting the yield of CD34+ cells in a collection are not yet fully understood. Throughout the literature, there has been conflicting evidence with regards to age, gender, disease status, and prior radiation. In addition to the factors that have already been explored, …

Gradient Directed Regularization For Sparse Gaussian Concentration Graphs, With Applications To Inference Of Genetic Networks, Hongzhe Li, Jiang Gui

UPenn Biostatistics Working Papers

Large-scale microarray gene expression data provide the possibility of constructing genetic networks or biological pathways. Gaussian graphical models have been suggested to provide an effective method for constructing such genetic networks. However, most of the available methods for constructing Gaussian graphs do not account for the sparsity of the networks and are computationally more demanding or infeasible, especially in the settings of high-dimension and low sample size. We introduce a threshold gradient descent regularization procedure for estimating the sparse precision matrix in the setting of Gaussian graphical models and demonstrate its application to identifying genetic networks. Such a procedure is …

Obesity, Self-Complexity, And Compartmentalization: On The Implications Of Obesity For Self-Concept Organization, Bruce E. Blaine, C. E. Johnson

Statistics Faculty/Staff Publications

The relationship between obesity and structural aspects of the self-concept was examined in adult women. Participants were 119 adult women [age range: 18-73, M=26.9; body mass index (BMI) range: 16.2-54.7, M=27.3] who completed measures of self-esteem, self-complexity, and the spontaneous self-concept. BMI was associated with less complex and more compartmentalized self-knowledge and more frequent mention of weight-stereotypic traits as self-descriptive. The findings are discussed in the context of research on obesity- related stigma.

Issues Of Processing And Multiple Testing Of Seldi-Tof Ms Proteomic Data, Merrill D. Birkner, Alan E. Hubbard, Mark J. Van Der Laan, Christine F. Skibola, Christine M. Hegedus, Martyn T. Smith

U.C. Berkeley Division of Biostatistics Working Paper Series

A new data filtering method for SELDI-TOF MS proteomic spectra data is described. We examined technical repeats (2 per subject) of intensity versus m/z (mass/charge) of bone marrow cell lysate for two groups of childhood leukemia patients: acute myeloid leukemia (AML) and acute lymphoblastic leukemia (ALL). As others have noted, the type of data processing as well as experimental variability can have a disproportionate impact on the list of "interesting" proteins (see Baggerly et al. (2004)). We propose a list of processing and multiple testing techniques to correct for 1) background drift; 2) filtering using smooth regression and cross-validated bandwidth …

Biomass Estimation Using Statistical And Neural Network Analysis Of Aster Data, Vijay O. Lulla

All-Inclusive List of Electronic Theses and Dissertations

This study assessed the performance of different biomass estimation methods using Advanced Spaceborne Thermal Emission and Reflection Radiometer (ASTER) lll data in a temperate forest. Multiple linear regression statistics of spectral band data and derived indices with biomass values were compared with advanced neural network models created with spectral band data and indices to model biomass values. Biomass for the study area was estimated using both of these methods and the results were discussed. The data were analyzed using multivariate statistical analysis. Correlation analysis and regression analysis were employed to understand the relationship of biomass to the spectral data and …

Distributed Blowing And Suction For The Purpose Of Streak Control In A Boundary Layer Subjected To A Favorable Pressure Gradient, Eric Forgoston, Anatoli Tumin, David E. Ashpis

Department of Applied Mathematics and Statistics Faculty Scholarship and Creative Works

An analysis of the optimal control by blowing and suction in order to generate streamwise velocity streaks is presented. The problem is examined using an iterative process that employs the Parabolized Stability Equations for an incompressible fluid along with its adjoint equations. In particular, distributions of blowing and suction are computed for both the normal and tangential velocity perturbations for various choices of parameters.

Three-Dimensional Wave Packet In A Hypersonic Boundary Layer, Eric Forgoston, Anatoli Tumin

Department of Applied Mathematics and Statistics Faculty Scholarship and Creative Works

A three-dimensional wave packet generated by a local disturbance in a hypersonic boundary layer flow is studied with the aid of the previously solved initial-value problem. The solution to this problem can be expanded in a biorthogonal eigenfunction system as a sum of discrete and continuous modes. A specific disturbance consisting of an initial temperature spot is considered, and the receptivity to this initial temperature spot is computed for both the two-dimensional and three-dimensional cases. Using previous analysis of the discrete and continuous spectrum, we numerically compute the inverse Fourier transform. The two-dimensional inverse Fourier transform is found for Mode …

The Stochastic Dance Of Early Hiv Infection, Stephen J. Merrill

Mathematics, Statistics and Computer Science Faculty Research and Publications

The stochastic nature of early HIV infection is described in a series of models, each of which captures aspects of the dance of HIV during the early stages of infection. It is to this highly variable target that the immune response must respond. The adaptability of the various components of the immune response is an important aspect of the system's operation, as the nature of the pathogens that the response will be required to respond to and the order in which those responses must be made cannot be known beforehand. As HIV infection has direct influence over cells responsible for …