Statistics and Probability Commons^™

A Scalable Supervised Subsemble Prediction Algorithm, Stephanie Sapp, Mark J. Van Der Laan

U.C. Berkeley Division of Biostatistics Working Paper Series

Subsemble is a flexible ensemble method that partitions a full data set into subsets of observations, fits the same algorithm on each subset, and uses a tailored form of V-fold cross-validation to construct a prediction function that combines the subset-specific fits with a second metalearner algorithm. Previous work studied the performance of Subsemble with subsets created randomly, and showed that these types of Subsembles often result in better prediction performance than the underlying algorithm fit just once on the full dataset. Since the final Subsemble estimator varies depending on the data used to create the subset-specific fits, different strategies for …

Multi-Tgdr, A Multi-Class Regularization Method, Identifies The Metabolic Profiles Of Hepatocellular Carcinoma And Cirrhosis Infected With Hepatitis B Or Hepatitis C Virus, Suyan Tian, Howard H. Chang, Chi Wang, Jing Jiang, Xiaomei Wang, Junqi Niu

Biostatistics Faculty Publications

BACKGROUND: Over the last decade, metabolomics has evolved into a mainstream enterprise utilized by many laboratories globally. Like other "omics" data, metabolomics data has the characteristics of a smaller sample size compared to the number of features evaluated. Thus the selection of an optimal subset of features with a supervised classifier is imperative. We extended an existing feature selection algorithm, threshold gradient descent regularization (TGDR), to handle multi-class classification of "omics" data, and proposed two such extensions referred to as multi-TGDR. Both multi-TGDR frameworks were used to analyze a metabolomics dataset that compares the metabolic profiles of hepatocellular carcinoma (HCC) …

Outcomes Of Recurrent Rectal Cancer After Transanal Excision, Sachin Vaid, Christopher Buzas, John S. Park Md, Robert J. Sinnott Md

Department of Surgery

No abstract provided.

A Bayesian Analysis Of The Spatial Concentration Of Individual Wealth In The Us North During The Nineteenth Century, Alice Kasakoff, Andrew Lawson, Emily Van Meter

Biostatistics Faculty Publications

Background: Kin effects can be difficult to distinguish from those of spatial proximity, since kin tend to live close to each other. Thus, past research showing correlations between the wealth of relatives may be showing the effects of proximity and shared locations, not the effects of kin.

Objective: What are the effects of kin and of spatial proximity upon wealth? This is studied both for fathers and sons and for brothers.

Methods: Data comes from a genealogical sample that has been linked to the US census of 1860. The genealogies allow us to identify fathers, sons, and …

Venue-Based Recruitment Of Women At Elevated Risk For Hiv: An Hiv Prevention Trials Network Study, Danielle F. Haley, Carol E. Golin, Wafaa M. El-Sadr, James P. Hughes, Jing Wang, Irene Kuo, +13 Additional Authors

Epidemiology Faculty Publications

Background: The challenge of identifying and recruiting U.S. women at elevated risk for HIV acquisition impedes prevention studies and services. HIV Prevention Trials Network (HPTN) 064 was a U.S. multisite, longitudinal cohort study designed to estimate HIV incidence among women living in communities with prevalent HIV and poverty. Venue-based sampling (VBS) methodologies and participant and venue characteristics are described.

Methods: Eligible women were recruited from 10 U.S. communities with prevalent HIV and poverty using VBS. Participant eligibility criteria included age 18–44 years, residing in a designated census tract/zip code, and self-report of at least one high-risk personal and/or male sexual …

Identification Of Slums In Mumbai, India: Unsupervised Classification Techniques, Frankie St. Amand

Thinking Matters Symposium Archive

Slums are contiguous settlements. Inhabitants lack access to safe water, sanitation and sewage infrastructure, secure housing tenure, uncrowded living space, and permanent, durable housing. Addressing these problematic trends begins with identifying contiguous settlements within Mumbai’s urban fabric. Classifications can be performed using satellite images and remote sensing techniques to yield accurate results. Through literature reviews, socio-cultural analysis, and examination of high resolution satellite imagery, this project aims to develop a systematic, accessible, and reproducible method of classifying Mumbai’s slums.

Building Footprint Extraction: A Land Use Classification Comparison Of Satellite Imagery Vs. Orthoimagery, Elisa Trepanier

Thinking Matters Symposium Archive

The Town of Windham, Maine Proposed Municipal Budget FY 2012-2013 called for $12,750 for professional services to develop a building footprint database. There are several municipal applications for building footprints including, but not limited to: Taxation/valuation of properties; setbacks for code enforcement; public safety planning; historical analysis of building patterns; general planning, where to encourage growth and redevelopment; zoning analysis and recreational planning. Due to the town’s in-house GIS (Geographic Information Systems) capabilities, the project will be developed within the Assessor’s Office and based on the 2012 Maine Office of GIS 6 Orthoimagery for Windham. The following is an effort …

Identifying High Crime Areas Using Spatial Analysis, Elisa Trepanier

Thinking Matters Symposium Archive

Crime incident locations and trends are examined spatially using GIS to produce maps that pinpoint high crime areas or Hot Spots Crime mapping aids police departments by identifying areas to allocate limited resources where and when they are most needed. This project introduces the availability of GIS technology to smaller police departments as a tool to assist in the development of crime prevention strategies. In this model crime incident reports for Windham, Maine are geocoded and patterns of motor vehicle and structure burglaries analyzed for date, time and location of incident. An addressing protocol is followed to protect victim privacy …

Effects Of Digital Elevation Model Resolution On Estimates Of Flood Damage In Westbrook, Maine, Scott Webber

Thinking Matters Symposium Archive

Accurate predictions of flood damage and economic losses are increasingly important as historically low-probability floods have become more common. Digital elevation models (DEMs) are essential to flood modeling as they define stream channel and floodplain morphology, the accuracy of which influence estimates of flood damage. The city of Westbrook, Maine experiences episodic flooding of the Presumpscot River; therefore, flooding in Westbrook was simulated with HAZUS, a model created by the Federal Emergency Management Agency. The model is designed to assess the costs associated with hazard scenarios including riverine flooding. HAZUS was run for a 100-year return period flood with both …

Section Abstracts: Statistics

Virginia Journal of Science

Abstracts of the Statistics Section for the 92nd Annual Meeting of the Virginia Academy of Science, May 13-15, 2014, Virginia Commonwealth University, Richmond, Virginia

Stream Crossing Barrier Prioritization Methods For Increasing Eastern Brook Trout Habitat In The Little Androscoggin River Watershed, Michele Windsor

Thinking Matters Symposium Archive

Eastern Brook Trout (Salvelinas fontanalis) are an important cold water fishery in the state of Maine. While populations in Maine are relatively abundant there has been decline in some parts of its range due in part to loss of habitat connectivity. Brook trout require access to specific types of stream habitat for spawning, feeding, and seasonal thermal refuges. Stream crossing structures such as undersized, poorly installed, or blocked culverts, as well as small remnant dams, can create barriers to accessing important stream habitat for brook trout. A recent Fish Barrier/Culvert Survey in the Little Androscoggin River Watershed provided data about …

Obesity Rates In Maine Exceed New England's Obesity Rate: A Look At Possible Contributing Factors, Susan Walker

Thinking Matters Symposium Archive

The purpose of this study was to explore adult obesity data to: Compare Maine’s rates with other New England states and explore differences in Maine counties by looking at known risk factors.

Hidden Trends In Nfl Data, Scott Santor

Statistics

This is an analysis on National Football League (NFL) data for the 2013-2014 regular season. The main goal is to find hidden trends in game data that can ultimately determine which factors are statistically significant to award a team with their ultimate objective, a win.

The main response variable to be examined is total wins throughout the regular season, and an alternative dependent variable is spread; the difference between a team’s points scored, and points against. Spread is analyzed to provide a different quantitative response variable that can be both positive and negative.

Game data was gathered from ESPN.com box …

The Impact Of Salsalate Treatment On Serum Levels Of Advanced Glycation End Products In Type 2 Diabetes., Joshua I Barzilay, Kathleen A. Jablonski, Vivian Fonseca, Steven E Shoelson, Allison B Goldfine, Christopher Strauch, Vincent M Monnier, Tinsal-T2d Research Consortium.

GW Biostatistics Center

OBJECTIVE Salsalate is a nonacetylated salicylate that lowers glucose levels in people with type 2 diabetes (T2D). Here we examined whether salsalate also lowered serum-protein-bound levels of early and advanced glycation end products (AGEs) that have been implicated in diabetic vascular complications. RESEARCH DESIGN AND METHODS Participants were from the Targeting Inflammation Using Salsalate for Type 2 Diabetes (TINSAL-T2D) study, which examined the impact of salsalate treatment on hemoglobin A1c (HbA1c) and a wide variety of other parameters. One hundred eighteen participants received salsalate, 3.5 g/day for 48 weeks, and 109 received placebo. Early glycation product levels (HbA1c and fructoselysine …

Genetic Risk Of Progression To Type 2 Diabetes And Response To Intensive Lifestyle Or Metformin In Prediabetic Women With And Without A History Of Gestational Diabetes Mellitus., Shannon D Sullivan, Kathleen A. Jablonski, Jose C Florez, Dana Dabelea, Paul W Franks, Sam Dagogo-Jack, Catherine Kim, William C Knowler, Costas A Christophi, Robert Ratner, Diabetes Prevention Program Research Group.

GW Biostatistics Center

OBJECTIVE The Diabetes Prevention Program (DPP) trial investigated rates of progression to diabetes among adults with prediabetes randomized to treatment with placebo, metformin, or intensive lifestyle intervention. Among women in the DPP, diabetes risk reduction with metformin was greater in women with prior gestational diabetes mellitus (GDM) compared with women without GDM but with one or more previous live births.

RESEARCH DESIGN AND METHODS We asked if genetic variability could account for these differences by comparing β-cell function and genetic risk scores (GRS), calculated from 34 diabetes-associated loci, between women with and without histories of GDM.

RESULTS β-Cell function was …

Metabolic Syndrome Components And Their Response To Lifestyle And Metformin Interventions Are Associated With Differences In Diabetes Risk In Persons With Impaired Glucose Tolerance, H. Florez, M. Temprosa, Trevor J. Orchard, Kieren J. Mather, Santica M. Marcovina, Elizabeth Barrett-Connor, Edward Horton, C. D. Saudek, F. X. Pi-Sunyer, R. E. Ratner, R. B. Goldberg

GW Biostatistics Center

Aims

To determine the association of metabolic syndrome (MetS) and its components with diabetes risk in participants with impaired glucose tolerance (IGT), and whether intervention-related changes in MetS lead to differences in diabetes incidence.

Methods

We used the National Cholesterol Education Program/Adult Treatment Panel III (NCEP/ATP III) revised MetS definition at baseline and intervention-related changes of its components to predict incident diabetes using Cox models in 3234 Diabetes Prevention Program (DPP) participants with IGT over an average follow-up of 3.2 years.

Results

In an intention-to-treat analysis, the demographic-adjusted hazard ratios (95% confidence interval) for diabetes in those with MetS (vs. …

What Residualizing Predictors In Regression Analyses Does (And What It Does Not Do), Lee H. Wurm, Sebastiano A. Fisicaro

Psychology Faculty Research Publications

Psycholinguists are making increasing use of regression analyses and mixed-effects modeling. In an attempt to deal with concerns about collinearity, a number of researchers orthogonalize predictor variables by residualizing (i.e., by regressing one predictor onto another, and using the residuals as a stand-in for the original predictor). In the current study, the effects of residualizing predictor variables are demonstrated and discussed using ordinary least-squares regression and mixed-effects models. Some of these effects are almost certainly not what the researcher intended and are probably highly undesirable. Most importantly, what residualizing does not do is change the result for the residualized variable, …

Pre-Maceration, Saignée And Temperature Affect Daily Evolution Of Pigment Extraction During Vinification, Ottorino L. Pantani, Federico M. Stefanini, Irene Lozzi, Luca Calamai, Alessandra Biondi Bartolini, Stefano Di Blasi

COBRA Preprint Series

Consumer demand for intensely coloured wines necessitates the systematic testing of pigment extraction in Sangiovese, a cultivar poor in easily extractable anthocyanins. Pre-fermentation (absent, cold soak pre-fermentation at 5 °C, cryomaceration by liquid N₂ addition), temperature (20 or 30 °C), and saignée were compared during vinification (800 kg). Concentrations of anthocyanins, non-anthocyanic flavonoids and SO₂-resistant pigments were recorded daily. A semiparametric Bayesian model permitted the kinetic description and the comparison of sigmoidal- and exponential-like curves. In total anthocyanins, saignée at 30 °C yielded a significant gain, later lost at drawing off; cryomaceration had little effect and cold …

Evaluating The Impact Of Genotype Errors On Rare Variant Tests Of Association, Kaitlyn Cook, Alejandra Benitez, Casey Fu, Nathan L. Tintle

Faculty Work Comprehensive List

The new class of rare variant tests has usually been evaluated assuming perfect genotype information. In reality, rare variant genotypes may be incorrect, and so rare variant tests should be robust to imperfect data. Errors and uncertainty in SNP genotyping are already known to dramatically impact statistical power for single marker tests on common variants and, in some cases, inflate the type I error rate. Recent results show that uncertainty in genotype calls derived from sequencing reads are dependent on several factors, including read depth, calling algorithm, number of alleles present in the sample, and the frequency at which an …

Highly Sensitive Noninvasive Cardiac Transplant Rejection Monitoring Using Targeted Quantification Of Donor-Specific Cell-Free Deoxyribonucleic Acid, Mats Hidestrand, Aoy Tomita-Mitchell, Pip M. Hidestrand, Arnold Oliphant, Mary Goetsch, Karl D. Stamm, Huan-Ling Liang, Chesney Castleberry, D. Woodrow Benson, Gail Stendahl, Pippa Simpson, Stuart Berger, James S. Tweddell, Steven Zangwill, Michael E. Mitchell

Mathematics, Statistics and Computer Science Faculty Research and Publications

No abstract provided.

Micro-Environment Causes Reversible Changes In Dna Methylation And Mrna Expression Profiles In Patient-Derived Glioma Stem Cells, Mehmet Baysan, Kevin Woolard, Serdar Bozdag, Gregory Riddick, Svetlana Kotliarova, Margaret C. Cam, Galina I. Belova, Susie Ahn, Wei Zhang, Hua Song, Jennifer Walling, Holly Stevenson, Paul Meltzer, Howard A. Fine

Mathematics, Statistics and Computer Science Faculty Research and Publications

In vitro and in vivo models are widely used in cancer research. Characterizing the similarities and differences between a patient's tumor and corresponding in vitro and in vivo models is important for understanding the potential clinical relevance of experimental data generated with these models. Towards this aim, we analyzed the genomic aberrations, DNA methylation and transcriptome profiles of five parental tumors and their matched in vitro isolated glioma stem cell (GSC) lines and xenografts generated from these same GSCs using high-resolution platforms. We observed that the methylation and transcriptome profiles of in vitro GSCs were significantly different from their corresponding …

Computational Pain Quantification And The Effects Of Age, Gender, Culture And Cause, Colin R. Ostberg

Master's Theses (2009 -)

Chronic pain affects more than 100 million Americans and more than 1.5 billion people worldwide. Pain is a multidimensional construct, expressed through a variety of means. Facial expressions are one such type of pain expression. Automatic facial expression recognition, and in particular pain expression recognition, are fields that have been studied extensively. However, nothing has explored the possibility of an automatic pain quantification algorithm, able to output pain levels based upon a facial image. Developed for a remote monitoring context, a computational pain quantification algorithm has been developed and validated by two distinct sets of data. The second set of …

Markov Chain Monte Carlo Bayesian Predictive Framework For Artificial Neural Network Committee Modeling And Simulation, Michael S. Goodrich

Computational Modeling & Simulation Engineering Theses & Dissertations

A logical inference method of properly weighting the outputs of an Artificial Neural Network Committee for predictive purposes using Markov Chain Monte Carlo simulation and Bayesian probability is proposed and demonstrated on machine learning data for non-linear regression, binary classification, and 1-of-k classification. Both deterministic and stochastic models are constructed to model the properties of the data. Prediction strategies are compared based on formal Bayesian predictive distribution modeling of the network committee output data and a stochastic estimation method based on the subtraction of determinism from the given data to achieve a stochastic residual using cross validation. Performance for Bayesian …

A Metaevaluation Of Energy Efficiency Evaluations, Brandy Brown

Dissertations

This study systematically reviews the methodological characteristics of energy efficiency evaluations and uses metaevaluation to assess its quality. Metaevaluation is used to systematically assess the quality of evaluation products, confirm that evaluations deliver sound findings and conclusions, are useful to the client, are credible, are ethically conducted, and are done as cost-effective as possible. The results of this study show that the ability to accurately assess evaluation for methodological quality using evaluations reports as a primary data source depends on the presence of detailed descriptions of evaluation methods. Furthermore, the study suggests that methodological variations of energy efficiency evaluations coalesce …

Multivariate Autoregressive Time Series Using Schweppe Weighted Wilcoxon Estimates, Jaime Burgos

Dissertations

The increasing needs of forecasting techniques has led to the popularity of the vector autoregressive model in multivariate time series analysis, which has become of typical use across different fields due to its simplicity in application. The traditional method for estimating the model parameters is the least squares minimization, due to the linear nature of the model and its similarity with multivariate linear regression. However, since least squares estimates are sensitive to outliers, more robust techniques have become of interest. This manuscript investigates a robust alternative by obtaining the estimates using a weighted Wilcoxon dispersion with Schweppe-type weights. The first …

A Metaevaluation Of Evaluations Of Health Care Programs That Employ The Chronic Care Model, Jan Fields

Dissertations

Background: The purpose of this dissertation is to explore the use of metaevaluation to evaluate the quality of healthcare studies conducted on programs that employ the Chronic Care Model (CCM) to provide chronic illness care. In this study, healthcare studies of CCM programs are regarded as program evaluations. Method: Using a non-experimental cross-sectional design, 28 healthcare studies of CCM programs were evaluated using the accuracy standards portion of the Program Evaluations Metaevaluation Checklist (Stufflebeam, 2011). The results of the metaevaluations were analyzed and compared to the HEAL grade of the same healthcare studies as determined by the Hierarchy of Evidence …

Using Multi-Objective Value Estimation To Support Predictive Analytics For Human Service Project Management, David D. Wingard

Dissertations

Human service organizations need outcome measurement approaches that support project management for efficiency and effectiveness. While, in recent years, human services have increased their capacity to manage data and measure outcomes empirically, several barriers remain. First, current outcome measurement practices are not designed to effectively support the management of human services programs for maximum efficiency and effectiveness. Second, human services organizations need a methodology to manage programs to identified outcomes. This dissertation explored meaningful solutions to both issues. In Paper 1 (Chapter II), this dissertation assessed strengths and limitations of current outcome evaluation approaches and suggested an innovative application of …

Improving The Design Of Cluster-Randomized Trials In Education: Informing The Selection Of Variance Design Parameter Values For Science Achievement Studies, Carl D. Westine

Dissertations

The purpose of this three-essay dissertation is to provide practical guidance to evaluators planning cluster-randomized trials (CRTs) of science achievement. In an educational setting, interventions are often administered at the cluster level, while outcomes are typically measured at the student level through standardized achievement testing. When evaluating an intervention, a CRT is appropriate because it allows for treatment to be modeled at a different level than the unit of analysis, and properly accounts for the violation of independence that occurs due to nesting. Accurately designing a CRT involves estimating variance parameters (i.e., intraclass correlations [ICCs] and percent of variance explained …

Harnessing Complexity: Analysis Methodology And Ethical Framework To Facilitate Utilization Of Video Data In Evaluations, Kurt A. Wilson

Dissertations

Most evaluations in the nonprofit and international development sectors are conducted in contexts of complexity; the specific intervention being evaluated is but one of many interrelated factors influencing the desired outcome. Video data, especially when directly generated by program participants, can provide both exceptionally rich qualitative data as well as contextually-relevant feedback within complex systems. Despite these unique strengths and opportunities, video data is underutilized in the field of evaluation. This dissertation addresses specific barriers associated with video data through three inter-related papers: Papers one and two (Chapters II and III) present the findings from two interrelated studies of an …

Bounds To Evaluate The Pure/Natural Direct Effect Without Cross-World Counterfactual Independence, Eric Tchetgen Tchetgen, Kelesitse Phiri

Harvard University Biostatistics Working Paper Series

No abstract provided.