Statistics and Probability Commons^™

A Dempster-Shafer Method For Multi-Sensor Fusion, Bethany G. Foley

Theses and Dissertations

The Dempster-Shafer Theory, a generalization of the Bayesian theory, is based on the idea of belief and as such can handle ignorance. When all of the required information is available, many data fusion methods provide a solid approach. Yet, most do not have a good way of dealing with ignorance. In the absence of information, these methods must then make assumptions about the sensor data. However, the real data may not fit well within the assumed model. Consequently, the results are often unsatisfactory and inconsistent. The Dempster-Shafer Theory is not hindered by incomplete models or by the lack of prior …

Breast Cancer Screening Practices Among American Indians And Alaska Natives In The Midwest, Christine M. Daley, Melissa Filippi, Aimee S. James, Sarah Brokenleg, Stacy Braiuca, K. A. Greiner, Won S. Choi

Journal of Health Disparities Research and Practice

American Indian and Alaska Native (AI/AN) women currently have some of the highest mortality rates from breast cancer for any racial/ethnic group in the United States and some of the lowest screening rates. However, current data are not available for regional differences in screening, which can result in dramatically different stage at diagnosis and mortality. We conducted surveys with 120 focus group participants in a needs assessment of mammography among AI/AN in the greater Kansas City metropolitan area and parts of Northeast Kansas. We found that among women under age 40, for whom recommended screenings include only annual clinical breast …

Race And Hepatitis C Management Within The Veterans Administration, Joahd Toure, Joshua Metlay, Sandford Schwartz, Knashawn Morales, David Kaplan, Peter Groeneveld

Journal of Health Disparities Research and Practice

Objective: To examine black-white differences in hepatitis C treatment within the Veterans Administration (VA) and determine whether racial variation in specialty consultation explains differences in hepatitis C treatment between blacks and whites.

Methods: We performed a retrospective cohort study of 1040 veterans meeting VA eligibility criteria for hepatitis C treatment. We used multiple imputation to handle missing race data. Specialty consultation was determined from the VA outpatient medical dataset and hepatitis C treatment was determined from the VA decision support system. Conditional logistic regression was used to examine the association between race and hepatitis C treatment as well as race …

Poverty And Mortality Disparities In Central Appalachia: Mountaintop Mining And Environmental Justice, Michael Hendryx

Journal of Health Disparities Research and Practice

Objectives. This study investigated the associations between poverty rates, Appalachian mountaintop coal mining, and age-adjusted total mortality rates to determine if persons exposed to this form of mining experience greater poverty and higher death rates compared to other types of mining or other areas of Appalachia.

Methods. Mortality rates, poverty rates, Appalachian designation and mining activity were examined for counties in Kentucky, Tennessee, Virginia and West Virginia (N=403). Linear least squares models tested for annual group differences from 2000-2007 in total and child poverty, and total mortality, based on mining type and Appalachian location. Nested linear models accounting for state-level …

Progestogens To Prevent Preterm Birth In Twin Pregnancies: An Individual Participant Data Meta-Analysis Of Randomized Trials., Ewoud Schuit, Sarah Stock, Rolf H.H. Groenwold, Kimberly Maurel, C. Andrew Combs, Thomas Garite, Catherine Y. Spong, Elizabeth A. Thom, Dwight J. Rouse, Steve N. Caritis, George Saade, Julia M. Zachary, Jane E. Norman, Line Rode, Katharina Klein, Ann Tabor, Elcin Cetingoz, John C. Morrison, Everett F. Magann, Christian M. Briery, Vicente Serra, Alfredo Perales, Juan Meseguer, Anwar Nassar, Arianne C. Lim, Karel G.M. Moons, Anneke Kwee, Ben W.J. Mol

Epidemiology Faculty Publications

Background

Preterm birth is the principal factor contributing to adverse outcomes in multiple pregnancies. Randomized controlled trials of progestogens to prevent preterm birth in twin pregnancies have shown no clear benefits. However, individual studies have not had sufficient power to evaluate potential benefits in women at particular high risk of early delivery (for example, women with a previous preterm birth or short cervix) or to determine adverse effects for rare outcomes such as intrauterine death.

Methods/design

We propose an individual participant data meta-analysis of high quality randomized, double-blind, placebo-controlled trials of progestogen treatment in women with a twin pregnancy. The …

Hierarchical Rank Aggregation With Applications To Nanotoxicology, Trina Patel, Donatello Telesca, Robert Rallo, Saji George, Xia Tian, Nel Andre

COBRA Preprint Series

The development of high throughput screening (HTS) assays in the field of nanotoxicology provide new opportunities for the hazard assessment and ranking of engineered nanomaterials (ENM). It is often necessary to rank lists of materials based on multiple risk assessment parameters, often aggregated across several measures of toxicity and possibly spanning an array of experimental platforms. Bayesian models coupled with the optimization of loss functions have been shown to provide an effective framework for conducting inference on ranks. In this article we present various loss function based ranking approaches for comparing ENM within experiments and toxicity parameters. Additionally, we propose …

Robustness Of Measures Of Interaction To Unmeasured Confounding, Eric J. Tchetgen Tchetgen, Tyler J. Vanderweele

Harvard University Biostatistics Working Paper Series

No abstract provided.

Bayesian Pollution Source Apportionment Incorporating Multiple Simultaneous Measurements, Jonathan Casey Christensen

Theses and Dissertations

We describe a method to estimate pollution profiles and contribution levels for distinct prominent pollution sources in a region based on daily pollutant concentration measurements from multiple measurement stations over a period of time. In an extension of existing work, we will estimate common source profiles but distinct contribution levels based on measurements from each station. In addition, we will explore the possibility of extending existing work to allow adjustments for synoptic regimes—large scale weather patterns which may effect the amount of pollution measured from individual sources as well as for particular pollutants. For both extensions we propose Bayesian methods …

Predicting Maximal Oxygen Consumption (Vo2max) Levels In Adolescents, Brent A. Shepherd

Theses and Dissertations

Maximal oxygen consumption (VO2max) is considered by many to be the best overall measure of an individual's cardiovascular health. Collecting the measurement, however, requires subjecting an individual to prolonged periods of intense exercise until their maximal level, the point at which their body uses no additional oxygen from the air despite increased exercise intensity, is reached. Collecting VO2max data also requires expensive equipment and great subject discomfort to get accurate results. Because of this inherent difficulty, it is often avoided despite its usefulness. In this research, we propose a set of Bayesian hierarchical models to predict VO2max levels in adolescents, …

Bootstrap-Based Inference On The Difference In The Means Of Two Correlated Functional Processes, Ciprian M. Crainiceanu, Ana-Maria Staicu, Shubankar Ray, Naresh Punjabi

Johns Hopkins University, Dept. of Biostatistics Working Papers

Nonparametric inference methods on the mean difference between two correlated Functional processes are proposed. We compare methods that: 1) incorporate different levels of smoothing of the mean and covariance; 2) preserve the sampling design; and 3) use parametric and nonparametric estimation of the mean functions. We apply our method to estimating the mean difference between average normalized δ-power of sleep electroencephalograms for 51 subjects with severe sleep apnea and 51 matched controls in the first 4 hours after sleep onset. Data are obtained from the Sleep Heart Health Study (SHHS), the largest community cohort study of sleep. While methods are …

Robustness Of Measures Of Interaction To Unmeasured Confounding, Eric J. Tchetgen Tchetgen, Tyler J. Vanderweele

COBRA Preprint Series

In this paper, we study the impact of unmeasured confounding on inference about a two-way interaction in a mean regression model with identity, log or logit link function. Necessary and sufficient conditions are established for a two-way interaction to be nonparametrically identified from the observed data, despite unmeasured confounding for the factors defining the interaction. A lung cancer data application illustrates the results.

The Effect Of Smoking On Tuberculosis Incidence In Burdened Countries, Natalie Noel Ellison

Theses and Dissertations

It is estimated that one third of the world's population is infected with tuberculosis. Though once thought a "dead" disease, tuberculosis is very much alive. The rise of drug resistant strains of tuberculosis, and TB-HIV coinfection have made tuberculosis an even greater worldwide threat. While HIV, poverty, and public health infrastructure are historically assumed to affect the burden of tuberculosis, recent research has been done to implicate smoking in this list. This analysis involves combining data from multiple sources in order determine if smoking is a statistically significant factor in predicting the number of incident tuberculosis cases in a country. …

Does Secondary Inflammatory Breast Cancer Represent Post-Surgical Metastatic Disease?, Salman Hashmi, Ladan Zolfaghari, Paul H. Levine

Epidemiology Faculty Publications

The phenomenon of accelerated tumor growth following surgery has been observed repeatedly and merits further study. Inflammatory breast carcinoma (IBC) is widely recognized as an extremely aggressive malignancy characterized by micrometastasis at the time of diagnosis, with one interesting subgroup defined as secondary IBC where pathologically identifiable IBC appears after surgical treatment of a primary non-inflammatory breast cancer. One possible mechanism can be related to the stimulation of dormant micrometastasis through local angiogenesis occurring as part of posttraumatic healing. In this report, we review cases of secondary IBC and others where localized trauma was followed by the appearance of IBC …

A Cautionary Note On Generalized Linear Models For Covariance Of Unbalanced Longitudinal Data, Jianhua Z. Huang, Min Chen, Mehdi Maadooliat, Mohsen Pourahmadi

Mathematics, Statistics and Computer Science Faculty Research and Publications

Missing data in longitudinal studies can create enormous challenges in data analysis when coupled with the positive-definiteness constraint on a covariance matrix. For complete balanced data, the Cholesky decomposition of a covariance matrix makes it possible to remove the positive-definiteness constraint and use a generalized linear model setup to jointly model the mean and covariance using covariates (Pourahmadi, 2000). However, this approach may not be directly applicable when the longitudinal data are unbalanced, as coherent regression models for the dependence across all times and subjects may not exist. Within the existing generalized linear model framework, we show how to overcome …

A Common Framework For Restriction Semigroups And Regular *-Semigroups, Peter R. Jones

Mathematics, Statistics and Computer Science Faculty Research and Publications

Left restriction semigroups have appeared at the convergence of several flows of research, including the theories of abstract semigroups, of partial mappings, of closure operations and even in logic. For instance, they model unary semigroups of partial mappings on a set, where the unary operation takes a map to the identity map on its domain. This perspective leads naturally to dual and two-sided versions of the restriction property. From a varietal perspective, these classes of semigroups–more generally, the corresponding classes of Ehresmann semigroups–derive from reducts of inverse semigroups, now taking a to a⁺=aa⁻¹ (or, dually, …

When 10 Trials Are Better Than 1000: An Evidentiary Perspective On Trial Sampling, Edward K. Cheng

Vanderbilt Law School Faculty Publications

In many mass tort cases, separately trying all individual claims is impractical, and thus a number of trial courts and commentators have explored the use of statistical sampling as a way of efficiently processing claims. Most discussions on the topic, however, implicitly assume that sampling is a “second best” solution: individual trials are preferred for accuracy, and sampling only justified under extraordinary circumstances. This Essay explores whether this assumption is really true. While intuitively one might think that individual trials would be more accurate at estimating liability than extrapolating from a subset of cases, the Essay offers three ways in …

Application Of Inter-Die Rank Statistics In Defect Detection, Vivek Bakshi

Dissertations and Theses

This thesis presents a statistical method to identify the test escapes. Test often acquires parametric measurements as a function of logical state of a chip. The usual method of classifying chips as pass or fail is to compare each state measurement to a test limit. Subtle manufacturing defects are escaping the test limits due to process variations in deep sub-micron technologies which results in mixing of healthy and faulty parametric test measurements. This thesis identifies the chips with subtle defects by using rank order of the parametric measurements. A hypothesis is developed that a defect is likely to disturb the …

On A Logistic Mixed Model Formulation Of A Quadratic Exponential Model For Correlated Binary Outcomes, Eric J. Tchetgen Tchetgen

Harvard University Biostatistics Working Paper Series

No abstract provided.

C2bat: A Novel Method For Association Between Ge- Netic Markers And Multiple Phenotypes, Melissa Naylor, Christoph Lange

Harvard University Biostatistics Working Paper Series

The purpose of this technical report is to describe a novel method developed to detect association between a genetic marker and multiple phenotypes. In order to obtain a one-degree of freedom test, a generalized principal component approach is suggested that aggregates the information about the genetic effect in the first prin- cipal component, while the remain principal components contain only environment noise. A limited simulation study is done validating the method. For scenarios in which the genetic effect is constant across all measurements and there is no envi- ronmental correlation between the measurements, preliminary results suggest that this method has …

Maximizing Network Lifetime On The Line With Adjustable Sensing Ranges, Amotz Bar-Noy, Ben Baumer

Statistical and Data Sciences: Faculty Publications

Given n sensors on a line, each of which is equipped with a unit battery charge and an adjustable sensing radius, what schedule will maximize the lifetime of a network that covers the entire line? Trivially, any reasonable algorithm is at least a 1/2-approximation, but we prove tighter bounds for several natural algorithms. We focus on developing a linear time algorithm that maximizes the expected lifetime under a random uniform model of sensor distribution. We demonstrate one such algorithm that achieves an average-case approximation ratio of almost 0.9. Most of the algorithms that we consider come from a family based …

Avoiding Boundary Estimates In Linear Mixed Models Through Weakly Informative Priors, Yeojin Chung, Sophia Rabe-Hesketh, Andrew Gelman, Jingchen Liu, Vincent Dorie

U.C. Berkeley Division of Biostatistics Working Paper Series

Variance parameters in mixed or multilevel models can be difficult to estimate, especially when the number of groups is small. We propose a maximum penalized likelihood approach which is equivalent to estimating variance parameters by their marginal posterior mode, given a weakly informative prior distribution. By choosing the prior from the gamma family with at least 1 degree of freedom, we ensure that the prior density is zero at the boundary and thus the marginal posterior mode of the group-level variance will be positive. The use of a weakly informative prior allows us to stabilize our estimates while remaining faithful …

On A Closed-Form Doubly Robust Estimator Of The Adjusted Odds Ratio For A Binary Exposure, Eric J. Tchetgen Tchetgen

Harvard University Biostatistics Working Paper Series

No abstract provided.

Formulae For Causal Mediation Analysis In An Odds Ratio Context Without A Normality Assumption For The Continuous Mediator, Eric J. Tchetgen Tchetgen

Harvard University Biostatistics Working Paper Series

No abstract provided.

A Cautionary Note On Specification Of The Correlation Structure In Inverse-Probability-Weighted Estimation For Repeated Measures, Eric J. Tchetgen Tchetgen, M. Maria Glymour, Jennifer Weuve, James Robins

Harvard University Biostatistics Working Paper Series

No abstract provided.

On Parametrization, Robustness And Sensitivity Analysis In A Marginal Structural Cox Proportional Hazards Model For Point Exposure, Eric J. Tchetgen Tchetgen, James M. Robins

Harvard University Biostatistics Working Paper Series

No abstract provided.

Multiple-Robust Estimation Of An Odds Ratio Interaction, Eric J. Tchetgen Tchetgen

Harvard University Biostatistics Working Paper Series

No abstract provided.

Inverse Odds Ratio-Weighted Estimation For Causal Mediation Analysis, Eric J. Tchetgen Tchetgen

Harvard University Biostatistics Working Paper Series

No abstract provided.

Cost-Effectiveness Of Reclassification Sampling For Prevalence Estimation, Airat Bekmetjev, Dirk Vanbruggen, Brian Mclellan, Benjamin Dewinkle, Eric Lunderberg, Nathan L. Tintle

Faculty Work Comprehensive List

Background: Typically, a two-phase (double) sampling strategy is employed when classifications are subject to error and there is a gold standard (perfect) classifier available. Two-phase sampling involves classifying the entire sample with an imperfect classifier, and a subset of the sample with the gold-standard.

Methodology/Principal Findings: In this paper we consider an alternative strategy termed reclassification sampling, which involves classifying individuals using the imperfect classifier more than one time. Estimates of sensitivity, specificity and prevalence are provided for reclassification sampling, when either one or two binary classifications of each individual using the imperfect classifier are available. Robustness of …

Effect Of Nutrition Changes On Foods Selected By Students In A Middle School-Based Diabetes Prevention Intervention Program: The Healthy Experience, Connie Mobley, Diane D. Stadler, Myrlene A. Staten, Laure El Ghormli, Bonnie Gillis, Jill Hartstein, Anna Maria Siega-Riz, Amy Virus, Healthy Study Group

GW Biostatistics Center

No abstract provided.

Repeat Prenatal Corticosteroid Prior To Preterm Birth: A Systematic Review And Individual Participant Data Meta-Analysis For The Precise Study Group (Prenatal Repeat Corticosteroid International Ipd Study Group: Assessing The Effects Using The Best Level Of Evidence) - Study Protocol, Caroline A. Crowther, Fariba Aghajafari, Lisa M. Askie, Elizabeth V. Asztalos, Peter Brocklehurst, Elizabeth A. Thom, +22 Additional Authors

GW Biostatistics Center

Background

The aim of this individual participant data (IPD) meta-analysis is to assess whether the effects of repeat prenatal corticosteroid treatment given to women at risk of preterm birth to benefit their babies are modified in a clinically meaningful way by factors related to the women or the trial protocol.

Methods/Design

The Prenatal Repeat Corticosteroid International IPD Study Group: assessing the effects using the best level of Evidence (PRECISE) Group will conduct an IPD meta-analysis. The PRECISE International Collaborative Group was formed in 2010 and data collection commenced in 2011. Eleven trials with up to 5,000 women and 6,000 infants …