Open Access. Powered by Scholars. Published by Universities.^{®}
Operations Research, Systems Engineering and Industrial Engineering Commons^{™}
Open Access. Powered by Scholars. Published by Universities.^{®}
 Keyword

 UARS distribution (2)
 MCMC (2)
 Graphical Models (2)
 Variance components (2)
 Instability (2)

 Department of Statistics (2)
 Degeneracy (2)
 Classification (2)
 Deep Learning (2)
 Orthogonal matrix (2)
 Community engagement (1)
 Bayesian statistics (1)
 Circular distribution Mixture Haar measure Jeffreys prior Markov chain Monte Carlo (1)
 Calcaneocuboid joint (1)
 BiasVariance tradeoff (1)
 Bayesian inference (1)
 Bilevel Optimization (1)
 Bayes (1)
 Credible interval (1)
 Closed loop supply chain (1)
 Credible intervals (1)
 Coverage rate (1)
 Control limits (1)
 CLT (1)
 Credible set of cones (1)
 Convergence rate (1)
 Blending (1)
 Bayes analysis (1)
 Circular distributions (1)
 Bayesian statistical methods (1)
 Publication Year
Articles 1  26 of 26
FullText Articles in Operations Research, Systems Engineering and Industrial Engineering
Forecasting Corn Yield With Machine Learning Ensembles, Mohsen Shahhosseini, Guiping Hu, Sotirios Archontoulis
Forecasting Corn Yield With Machine Learning Ensembles, Mohsen Shahhosseini, Guiping Hu, Sotirios Archontoulis
Industrial and Manufacturing Systems Engineering Publications
The emerge of new technologies to synthesize and analyze big data with highperformance computing, has increased our capacity to more accurately predict crop yields. Recent research has shown that Machine learning (ML) can provide reasonable predictions, faster, and with higher flexibility compared to simulation crop modeling. The earlier the prediction during the growing season the better, but this has not been thoroughly investigated as previous studies considered all data available to predict yields. This paper provides a machine learning based framework to forecast corn yields in three US Corn Belt states (Illinois, Indiana, and Iowa) considering complete and partial inseason ...
Optimizing Ensemble Weights And Hyperparameters Of Machine Learning Models For Regression Problems, Mohsen Shahhosseini, Guiping Hu, Hieu Pham
Optimizing Ensemble Weights And Hyperparameters Of Machine Learning Models For Regression Problems, Mohsen Shahhosseini, Guiping Hu, Hieu Pham
Industrial and Manufacturing Systems Engineering Publications
Aggregating multiple learners through an ensemble of models aims to make better predictions by capturing the underlying distribution more accurately. Different ensembling methods, such as bagging, boosting and stacking/blending, have been studied and adopted extensively in research and practice. While bagging and boosting intend to reduce variance and bias, respectively, blending approaches target both by finding the optimal way to combine base learners to find the best tradeoff between bias and variance. In blending, ensembles are created from weighted averages of multiple base learners. In this study, a systematic approach is proposed to find the optimal weights to create ...
Computational Aspects Of Bayesian Solution Estimators In Stochastic Optimization, Danial Davarnia, Burak Kocuk, Gerard Cornuejols
Computational Aspects Of Bayesian Solution Estimators In Stochastic Optimization, Danial Davarnia, Burak Kocuk, Gerard Cornuejols
Industrial and Manufacturing Systems Engineering Publications
We study a class of stochastic programs where some of the elements in the objective function are random, and their probability distribution has unknown parameters. The goal is to ﬁnd a good estimate for the optimal solution of the stochastic program using data sampled from the distribution of the random elements. We investigate two common optimization criteria for evaluating the quality of a solution estimator, one based on the diﬀerence in objective values, and the other based on the Euclidean distance between solutions. We use risk as the expected value of such criteria over the sample space. Under a Bayesian ...
Biclustermd: An R Package For Biclustering With Missing Values, John Reisner, Hieu Pham, Sigurdur Olafsson, Stephen B. Vardeman, Jing Li
Biclustermd: An R Package For Biclustering With Missing Values, John Reisner, Hieu Pham, Sigurdur Olafsson, Stephen B. Vardeman, Jing Li
Industrial and Manufacturing Systems Engineering Publications
Biclustering is a statistical learning technique that attempts to find homogeneous partitions of rows and columns of a data matrix. For example, movie ratings might be biclustered to group both raters and movies. biclust is a current R package allowing users to implement a variety of biclustering algorithms. However, its algorithms do not allow the data matrix to have missing values. We provide a new R package, biclustermd, which allows users to perform biclustering on numeric data even in the presence of missing values.
A Bayesian StateSpace Model Using AgeAtHarvest Data For Estimating The Population Of Black Bears (Ursus Americanus) In Wisconsin, Maximilian L. Allen, Andrew S. Norton, Glenn Stauffer, Nathan M. Roberts, Yanshi Luo, Qing Li, David Macfarland, Timothy R. Van Deelen
A Bayesian StateSpace Model Using AgeAtHarvest Data For Estimating The Population Of Black Bears (Ursus Americanus) In Wisconsin, Maximilian L. Allen, Andrew S. Norton, Glenn Stauffer, Nathan M. Roberts, Yanshi Luo, Qing Li, David Macfarland, Timothy R. Van Deelen
Industrial and Manufacturing Systems Engineering Publications
Population estimation is essential for the conservation and management of fish and wildlife, but accurate estimates are often difficult or expensive to obtain for cryptic species across large geographical scales. Accurate statistical models with manageable financial costs and field efforts are needed for hunted populations and using ageatharvest data may be the most practical foundation for these models. Several rigorous statistical approaches that use ageatharvest and other data to accurately estimate populations have recently been developed, but these are often dependent on (a) accurate prior knowledge about demographic parameters of the population, (b) auxiliary data, and (c) initial population size ...
The Study Design Elements Employed By Researchers In Preclinical Animal Experiments From Two Research Domains And Implications For Automation Of Systematic Reviews, Annette M. O'Connor, Sarah C. Totton, Jonah C. Cullen, Mahmood Ramezani, Vijay Kalivarapu, Chaohui Yuan, Stephen B. Gilbert
The Study Design Elements Employed By Researchers In Preclinical Animal Experiments From Two Research Domains And Implications For Automation Of Systematic Reviews, Annette M. O'Connor, Sarah C. Totton, Jonah C. Cullen, Mahmood Ramezani, Vijay Kalivarapu, Chaohui Yuan, Stephen B. Gilbert
Veterinary Diagnostic and Production Animal Medicine Publications
Systematic reviews are increasingly using data from preclinical animal experiments in evidence networks. Further, there are everincreasing efforts to automate aspects of the systematic review process. When assessing systematic bias and unitofanalysis errors in preclinical experiments, it is critical to understand the study design elements employed by investigators. Such information can also inform prioritization of automation efforts that allow the identification of the most common issues. The aim of this study was to identify the design elements used by investigators in preclinical research in order to inform unique aspects of assessment of bias and error in preclinical research. Using 100 ...
Properties And Bayesian Fitting Of Restricted Boltzmann Machines, Andee Kaplan, Daniel J. Nordman, Stephen B. Vardeman
Properties And Bayesian Fitting Of Restricted Boltzmann Machines, Andee Kaplan, Daniel J. Nordman, Stephen B. Vardeman
Statistics Preprints
A restricted Boltzmann machine (RBM) is an undirected graphical model constructed for discrete or continuous random variables, with two layers, one hidden and one visible, and no conditional dependency within a layer. In recent years, RBMs have risen to prominence due to their connection to deep learning. By treating a hidden layer of one RBM as the visible layer in a second RBM, a deep architecture can be created. RBMs are thought to thereby have the ability to encode very complex and rich structures in data, making them attractive for supervised learning. However, the generative behavior of RBMs is largely ...
Estimating A ServiceLife Distribution Based On Production Counts And A Failure Database, Kenneth J. Ryan, Michael S. Hamada, Stephen B. Vardeman
Estimating A ServiceLife Distribution Based On Production Counts And A Failure Database, Kenneth J. Ryan, Michael S. Hamada, Stephen B. Vardeman
Statistics Publications
Problem: A manufacturer wanted to compare the servicelife distributions of two similar products. These concern product lifetimes after installation (not manufacture). For each product, there were available production counts and an imperfect database providing information on failing units. In the real case, these units were expensive repairable units warrantied against repairs. Failure (of interest here) was relatively rare and driven by a different mode/mechanism than ordinary repair events (not of interest here).
Approach: Data models for the service life based on a standard parametric lifetime distribution and a related limited failure population were developed. These models were used to ...
On The Instability And Degeneracy Of Deep Learning Models, Andee Kaplan, Daniel J. Nordman, Stephen B. Vardeman
On The Instability And Degeneracy Of Deep Learning Models, Andee Kaplan, Daniel J. Nordman, Stephen B. Vardeman
Statistics Preprints
A probability model exhibits instability if small changes in a data outcome result in large, and often unanticipated, changes in probability. This instability is a property of the probability model, rather than the fitted parameter vector. For correlated data structures found in several application areas, there is increasing interest in predicting/identifying such sensitivity in model probability structure. We consider the problem of quantifying instability for general probability models defined on sequences of observations, where each sequence of length N has a finite number of possible values. A sequence of probability models results, indexed by N, that accommodates data of ...
Methodologies For Studying HumanMicroclimate Interactions For Resilient, Smart City DecisionMaking, Ulrike Passe, Nadia Anderson, Kris De Brabanter, Michael C. Dorneich, Caroline Krejci, Alenka Poplin, Linda Shenk
Methodologies For Studying HumanMicroclimate Interactions For Resilient, Smart City DecisionMaking, Ulrike Passe, Nadia Anderson, Kris De Brabanter, Michael C. Dorneich, Caroline Krejci, Alenka Poplin, Linda Shenk
Industrial and Manufacturing Systems Engineering Conference Proceedings and Posters
Creating sustainable, resilient cities requires integrating an understanding of human behavior and decisionmaking about the built environment within an expanding range of spatial, political, and cultural contexts. Resilience—the ability to survive from and adapt to extreme or sudden stresses—emphasizes the importance of participation by a broad range of stakeholders in making decisions for the future. Smart cities leverage technology and data collected from the community and its stakeholders to inform and support these decisions. Energy usage in cities starts with people interacting with their environments, such as occupants interacting with the buildings in which they live and work ...
Modern Measurement, Probability, And Statistics: Some Generalities And Multivariate Illustrations, Stephen B. Vardeman
Modern Measurement, Probability, And Statistics: Some Generalities And Multivariate Illustrations, Stephen B. Vardeman
Industrial and Manufacturing Systems Engineering Publications
In broad terms, effective probability modeling of modern measurement requires the development of (usually parametric) distributions for increasingly complex multivariate outcomes driven by the physical realities of particular measurement technologies. “Differences” between measures of distribution center and truth function as “bias.” Model features that allow hierarchical compounding of variation function to describe “variance components” like “repeatability,” “reproducibility,” “batchtobatch variation,” etc. Mixture features in models allow for description (and subsequent downweighting) of outliers. For a variety of reasons (including highdimensionality of parameter spaces relative to typical sample sizes, the ability to directly include “Type B” considerations in assessing uncertainty, and the ...
A PseudoLikelihood Analysis For Incomplete Warranty Data With A Time Usage Rate Variable And Production Counts, Yu Qiu, Danial J. Nordman, Stephen B. Vardeman
A PseudoLikelihood Analysis For Incomplete Warranty Data With A Time Usage Rate Variable And Production Counts, Yu Qiu, Danial J. Nordman, Stephen B. Vardeman
Industrial and Manufacturing Systems Engineering Publications
The most direct purpose of collecting warranty data is tracking associated costs. However, they are also useful for quantifying a relationship between use rate and product timetofirstfailure and for estimating the distribution of product timetofirstfailure (which is modeled in this article as depending on use rate and a unit potential life length under continuous use). Employing warranty data for such reliability analysis purposes is typically complicated by the fact that some parts of some warranty data records are missing. A pseudolikelihood methodology is introduced to deal with some kinds of incomplete warranty data (such as that available in a motivating ...
A Wrapped Trivariate Normal Distribution And Bayes Inference For 3D Rotations, Yu Qiu, Daniel J. Nordman, Stephen B. Vardeman
A Wrapped Trivariate Normal Distribution And Bayes Inference For 3D Rotations, Yu Qiu, Daniel J. Nordman, Stephen B. Vardeman
Industrial and Manufacturing Systems Engineering Publications
For modeling orientation data represented as 3 × 3 rotation matrices, we develop a wrapped trivariate normal distribution (wTND) under which random rotations have simple geometric construction as symmetric errors about a mean. While of interest in its own right, the wTND also provides simple and effective approximations to the isotropic Gaussian distribution on rotations, with some advantages over approximations based on other commonly used models. We develop noninformative Bayes inference for the wTND via Markov Chain Monte Carlo methods that allow straightforward computations in a model where maximum likelihood is undefined. Credible regions for model parameters (including a fixed 3 ...
OneSample Bayes Inference For Symmetric Distributions Of 3D Rotations, Yu Qiu, Danial J. Nordman, Stephen B. Vardeman
OneSample Bayes Inference For Symmetric Distributions Of 3D Rotations, Yu Qiu, Danial J. Nordman, Stephen B. Vardeman
Industrial and Manufacturing Systems Engineering Publications
A variety of existing symmetric parametric models for 3D rotations found in both statistical and materials science literatures are considered from the point of view of the “uniformaxisrandomspin” (UARS) construction. Onesample Bayes methods for noninformative priors are provided for all of these models and attractive frequentist properties for corresponding Bayes inference on the model parameters are confirmed. Taken together with earlier work, the broad efficacy of noninformative Bayes inference for symmetric distributions on 3D rotations is conclusively demonstrated.
Majority Voting By Independent Classifiers Can Increase Error Rates, Stephen B. Vardeman, Max Morris
Majority Voting By Independent Classifiers Can Increase Error Rates, Stephen B. Vardeman, Max Morris
Industrial and Manufacturing Systems Engineering Publications
The technique of “majority voting” of classifiers is used in machine learning with the aim of constructing a new combined classification rule that has better characteristics than any of a given set of rules. The “Condorcet Jury Theorem” is often cited, incorrectly, as support for a claim that this practice leads to an improved classifier (i.e., one with smaller error probabilities) when the given classifiers are sufficiently good and are uncorrelated. We specifically address the case of twocategory classification, and argue that a correct claim can be made for independent (not just uncorrelated) classification errors (not the classifiers themselves ...
Bayes Inference For A Tractable New Class Of NonSymmetric Distributions For 3Dimensional Rotations, Melissa Ann Bingham, Danial J. Nordman, Stephen B. Vardeman
Bayes Inference For A Tractable New Class Of NonSymmetric Distributions For 3Dimensional Rotations, Melissa Ann Bingham, Danial J. Nordman, Stephen B. Vardeman
Industrial and Manufacturing Systems Engineering Publications
Both existing models for nonsymmetric distributions on 3dimensional rotations and their associated onesample inference methods have serious limitations in terms of both interpretability and ease of use. Based on the intuitively appealing Uniform Axis Random Spin (UARS) construction of Bingham, Nordman, and Vardeman (2009) for symmetric families of distributions, we propose new highly interpretable and tractable classes of nonsymmetric distributions that are derived from mixing UARS distributions. These have an appealing Preferred AxisRandom Spin (PARS) construction and (unlike existing models) directly interpretable parameters. Noninformative onesample Bayes inference in these models is a direct generalization of UARS methods introduced in Bingham ...
Heavy Traffic Analysis Of A Simple ClosedLoop Supply Chain, Arka P. Ghosh, Sarah M. Ryan, Lizhi Wang, Ananda Weerasinghe
Heavy Traffic Analysis Of A Simple ClosedLoop Supply Chain, Arka P. Ghosh, Sarah M. Ryan, Lizhi Wang, Ananda Weerasinghe
Industrial and Manufacturing Systems Engineering Publications
We consider a closedloop supply chain where new products are produced to order and returned products are refurbished for reselling. The solution to apricesetting problem enforces the heavy traffic condition, under which we address the production rate control problem for two types of cost functions. Wesolve a driftcontrol problem for an approximate system driven by a correlated twodimensional Brownian motion. The solutions to this system are then used to obtain asymptotically optimal control policies. We also conduct a numerical study to explore the effects of different parameters on the optimal production rates and the resulting costs.
Bayes OneSample And OneWay Random Effects Analyses For 3D Orientations With Application To Materials Science, Melissa Ann Bingham, Stephen B. Vardeman, Daniel J. Nordman
Bayes OneSample And OneWay Random Effects Analyses For 3D Orientations With Application To Materials Science, Melissa Ann Bingham, Stephen B. Vardeman, Daniel J. Nordman
Industrial and Manufacturing Systems Engineering Publications
We consider Bayes inference for a class of distributions on orientations in 3 dimensions described by 3×3 rotation matrices. Noninformative priors are identified and MetropolisHastings within Gibbs algorithms are used to generate samples from posterior distributions in onesample and oneway random effects models. A simulation study investigates the performance of Bayes analyses based on noninformative priors in the onesample case, making comparisons to quasilikelihood inference. A second simulation study investigates the behavior of posteriors for some informative priors. Bayes oneway random effect analyses of orientation matrix data are then developed and the Bayes methods are illustrated in a materials ...
Modeling And Inference For Measured Crystal Orientations And A Tractable Class Of Symmetric Distributions For Rotations In Three Dimensions, Melissa Ann Bingham, Daniel J. Nordman, Stephen B. Vardeman
Modeling And Inference For Measured Crystal Orientations And A Tractable Class Of Symmetric Distributions For Rotations In Three Dimensions, Melissa Ann Bingham, Daniel J. Nordman, Stephen B. Vardeman
Industrial and Manufacturing Systems Engineering Publications
Electron backscatter diffraction (EBSD) is a technique used in materials science to study the microtexture of metals, producing data that measure the orientations of crystals in a specimen. We examine the precision of such data based on a useful class of distributions on orientations in three dimensions (as represented by 3×3 orthogonal matrices with positive determinants). Although such modeling has received attention in the statistical literature, the approach taken typically has been based on general “special manifold” considerations, and the resulting methodology may not be easily accessible to nonspecialists. We take a more direct modeling approach, beginning from a ...
Uniformly HyperEfficient Bayes Inference In A Class Of Nonregular Problems, Danial J. Nordman, Stephen B. Vardeman, Melissa Ann Bingham
Uniformly HyperEfficient Bayes Inference In A Class Of Nonregular Problems, Danial J. Nordman, Stephen B. Vardeman, Melissa Ann Bingham
Industrial and Manufacturing Systems Engineering Publications
We present a tractable class of nonregular continuous statistical models where 1) likelihoods have multiple singularities and ordinary maximum likelihood is intrinsically unavailable, but 2) Bayes procedures achieve convergence rates better than n−1 across the whole parameter space. In fact, for every p>1, there is a member of the class for which the posterior distribution is consistent at rate n−puniformly in the parameter.
Calibration, Error Analysis, And Ongoing Measurement Process Monitoring For Mass Spectrometry, Stephen B. Vardeman, Joanne Wendelberger, Lily Wang
Calibration, Error Analysis, And Ongoing Measurement Process Monitoring For Mass Spectrometry, Stephen B. Vardeman, Joanne Wendelberger, Lily Wang
Industrial and Manufacturing Systems Engineering Publications
We consider problems of quantifying and monitoring accuracy and precision of measurement in mass spectrometry, particularly in contexts where there is unavoidable daytoday/periodtoperiod changes in instrument sensitivity. First, we consider the issue of estimating instrument sensitivity based on data from a typical calibration study. Simple methodofmoments methods, likelihoodbased methods, and Bayes methods based on the oneway random effects model are illustrated. Then, we consider subsequently assessing the precision of an estimate of a mole fraction of a gas of interest in an unknown. Finally, we turn to the problem of ongoing measurement process monitoring and illustrate appropriate setup of ...
LikelihoodBased Statistical Estimation From Quantized Data, Stephen B. Vardeman, ChiangSheng Lee
LikelihoodBased Statistical Estimation From Quantized Data, Stephen B. Vardeman, ChiangSheng Lee
Industrial and Manufacturing Systems Engineering Publications
Most standard statistical methods treat numerical data as if they were real (infinitenumberofdecimalplaces) observations. The issue of quantization or digital resolution can render such methods inappropriate and misleading. This article discusses some of the difficulties of interpretation and corresponding difficulties of inference arising in even very simple measurement contexts, once the presence of quantization is admitted. It then argues (using the simple case of confidence interval estimation based on a quantized random sample from a normal distribution as a vehicle) for the use of statistical methods based on "rounded data likelihood functions" as an effective way of handling the matter.
Sheppard's Correction For Variances And The "Quantization Noise Model", Stephen B. Vardeman
Sheppard's Correction For Variances And The "Quantization Noise Model", Stephen B. Vardeman
Industrial and Manufacturing Systems Engineering Publications
In this paper, we examine the relevance of Sheppard's correction for variances and (both the original and a valid weak form of) the socalled "quantization noise model" to understanding the effects of integer rounding on continuous random variables. We further consider whether there is any real relationship between the two. We observe that the strong form of the model is not really relevant to describing rounding effects. We demonstrate using simple cases the substantial limitations of the Sheppard correction, and use simple versions of a weak form of the model to establish that there is no real connection between ...
Likelihood And Bayesian Methods For Accurate Identification Of Measurement Biases In Pseudo SteadyState Processes, Sriram Devanathan, Stephen B. Vardeman, Derrick K. Rollins Sr.
Likelihood And Bayesian Methods For Accurate Identification Of Measurement Biases In Pseudo SteadyState Processes, Sriram Devanathan, Stephen B. Vardeman, Derrick K. Rollins Sr.
Industrial and Manufacturing Systems Engineering Publications
Two new approaches are presented for improved identification of measurement biases in linear pseudo steadystate processes. Both are designed to detect a change in the mean of a measured variable leading to an inference regarding the presence of a biased measurement. The first method is based on a likelihood ratio test for the presence of a mean shift. The second is based on a Bayesian decision rule (relying on prior distributions for unknown parameters) for the detection of a mean shift. The performance of these two methods is compared with that of a method given by Devanathan et al. (2000 ...
Development Programs For OneShot Systems Using MultipleState Design Reliability Models, Suntichai Shevasuthisilp, Stephen B. Vardeman
Development Programs For OneShot Systems Using MultipleState Design Reliability Models, Suntichai Shevasuthisilp, Stephen B. Vardeman
Industrial and Manufacturing Systems Engineering Publications
Design reliability at the beginning of a product development program is typically low, and development costs can account for a large proportion of total product cost. We consider how to conduct development programs (series of tests and redesigns) for oneshot systems (which are destroyed at first use or during testing). In rough terms, our aim is to both achieve high final design reliability and spend as little of a fixed budget as possible on development. We employ multiplestate reliability models. Dynamic programming is used to identify a best testandredesign strategy and is shown to be presently computationally feasible for at ...
Form Error Estimation Using Spatial Statistics, TaiHung Yang, John K. Jackman
Form Error Estimation Using Spatial Statistics, TaiHung Yang, John K. Jackman
Industrial and Manufacturing Systems Engineering Publications
Form error estimation is an essential step in the assessment of product geometry created through one or more manufacturing processes. We present a new method using spatial statistics to estimate form error. Using large sets of uniform sample points measured from five common machined surfaces, we compare the form error estimates using individual points and fitted surfaces obtained through spatial statistical methods. The results show that spatial statistics can provide more accurate estimates of form error under certain conditions.