BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//UNC Statistics & Operational Research - ECPv5.6.0//NONSGML v1.0//EN
CALSCALE:GREGORIAN
METHOD:PUBLISH
X-WR-CALNAME:UNC Statistics & Operational Research
X-ORIGINAL-URL:https://stor.unc.edu
X-WR-CALDESC:Events for UNC Statistics & Operational Research
BEGIN:VTIMEZONE
TZID:America/Halifax
BEGIN:DAYLIGHT
TZOFFSETFROM:-0400
TZOFFSETTO:-0300
TZNAME:ADT
DTSTART:20190310T060000
END:DAYLIGHT
BEGIN:STANDARD
TZOFFSETFROM:-0300
TZOFFSETTO:-0400
TZNAME:AST
DTSTART:20191103T050000
END:STANDARD
TZID:America/New_York
BEGIN:DAYLIGHT
TZOFFSETFROM:-0500
TZOFFSETTO:-0400
TZNAME:EDT
DTSTART:20190310T070000
END:DAYLIGHT
BEGIN:STANDARD
TZOFFSETFROM:-0400
TZOFFSETTO:-0500
TZNAME:EST
DTSTART:20191103T060000
END:STANDARD
BEGIN:DAYLIGHT
TZOFFSETFROM:-0500
TZOFFSETTO:-0400
TZNAME:EDT
DTSTART:20200308T070000
END:DAYLIGHT
BEGIN:STANDARD
TZOFFSETFROM:-0400
TZOFFSETTO:-0500
TZNAME:EST
DTSTART:20201101T060000
END:STANDARD
TZID:America/New_York
BEGIN:DAYLIGHT
TZOFFSETFROM:-0500
TZOFFSETTO:-0400
TZNAME:EDT
DTSTART:20190310T070000
END:DAYLIGHT
BEGIN:STANDARD
TZOFFSETFROM:-0400
TZOFFSETTO:-0500
TZNAME:EST
DTSTART:20191103T060000
END:STANDARD
END:VTIMEZONE
BEGIN:VEVENT
DTSTART;TZID=America/Halifax:20190805T090000
DTEND;TZID=America/Halifax:20190805T110000
DTSTAMP:20211208T010014
CREATED:20190718T094128Z
LAST-MODIFIED:20190718T094128Z
UID:9866-1564995600-1565002800@stor.unc.edu
SUMMARY:Boot Camp - Analysis
DESCRIPTION:
URL:https://stor.unc.edu/event/boot-camp-analysis/2019-08-05/
END:VEVENT
BEGIN:VEVENT
DTSTART;TZID=America/Halifax:20190805T130000
DTEND;TZID=America/Halifax:20190805T150000
DTSTAMP:20211208T010014
CREATED:20190718T094517Z
LAST-MODIFIED:20190718T094517Z
UID:9874-1565010000-1565017200@stor.unc.edu
SUMMARY:Boot Camp - Linear Algebra
DESCRIPTION:
URL:https://stor.unc.edu/event/boot-camp-linear-algebra/2019-08-05/
LOCATION:125 Hanes Hall\, Hanes Hall\, Chapel Hill\, NC\, 27599\, United States
END:VEVENT
BEGIN:VEVENT
DTSTART;TZID=America/Halifax:20190812T090000
DTEND;TZID=America/Halifax:20190812T130000
DTSTAMP:20211208T010014
CREATED:20190321T084148Z
LAST-MODIFIED:20190321T084148Z
UID:9851-1565600400-1565614800@stor.unc.edu
SUMMARY:CWE exams
DESCRIPTION:Today’s exam: 612 & 614
URL:https://stor.unc.edu/event/cwe-exams-6/
LOCATION:107 Hanes Hall\, Hanes Hall\, Chapel Hill\, NC\, 27599\, United States
END:VEVENT
BEGIN:VEVENT
DTSTART;TZID=America/Halifax:20190812T130000
DTEND;TZID=America/Halifax:20190812T150000
DTSTAMP:20211208T010014
CREATED:20190809T105418Z
LAST-MODIFIED:20190809T105418Z
UID:9884-1565614800-1565622000@stor.unc.edu
SUMMARY:Ph.D. Defense: Yichen Tu
DESCRIPTION:The Department of \nStatistics and Operations Research \nThe University of North Carolina at Chapel Hill \nYichen Tu\nQueuing Systems with Strategic and Learning Customers \n(Under the direction of Serhan Ziya and Nur Sunar) \n\nIn many service systems customers are strategic and can make their own decisions. In particular\, customers can be delay-sensitive and they will leave the system if they think the waiting time is too long. For the service provider\, it is important to understand customers’ behaviors and choose the appropriate system design. This dissertation consists of two research projects. \n\nThe first project studies the pooling decision when customers are strategic. It is generally accepted that operating with a combined (i.e.\, pooled) queue rather than separate (i.e.\, dedicated) queues is beneficial mainly because pooling queues reduces long-run average sojourn time. In fact\, this is a well-established result in the literature when jobs cannot make decisions and servers and jobs are identical. An important corollary of this finding is that pooling queues improves social welfare in the aforementioned setting. We consider an observable multi-server queueing system which can be operated with either dedicated queues or a pooled one. Customers are delay-sensitive and they decide to join or balk based on queue length information upon arrival. In this setting\, we prove that\, contrary to the common understanding\, pooling queues can considerably increase the long-run average sojourn time so that the pooled system results in strictly smaller social welfare (and strictly smaller consumer surplus) than the dedicated system under certain conditions. Specifically\, pooling queues leads to performance loss when the arrival-rate-to-service-rate ratio and the relative benefit of service are both large. We also prove that performance loss due to pooling queues can be significant. Our numerical studies demonstrate that pooling queues can decrease the social welfare (and the consumer surplus) by more than 95%. The benefit of pooling is commonly believed to increase with the system size. In contrast to this belief\, our analysis shows that when delay-sensitive customers make rational joining decisions\, the magnitude of the performance loss due to pooling can strictly increase with the system size. \n\nThe second project studies the learning behavior when customers don’t have full information of the service speed. We consider a single-server queueing system where customers make join- ing and abandonment decisions when the service rate is unknown. We study different ways in which customers process service-related information\, and how these impact the performance of a service provider. Specifically\, we analyze forward-looking\, myopic and naive information process- ing behaviors by customers. Forward-looking customers learn about the service rate in a Bayesian framework by using their observations while waiting in the queue. Moreover\, they make their abandonment decisions considering both belief and potential future payoffs. On the other hand\, naive customers ignore the available information when they make their decisions. We prove that regardless of the way in which the information is processed by customers\, a customer’s optimal joining and abandonment policy is of threshold-type. There is a rich literature that establishes that forward-looking customers are detrimental to a firm in settings different than queueing. In contrast to this common understanding\, we prove that for service systems\, forward-looking customers are beneficial to the firm under certain conditions.
URL:https://stor.unc.edu/event/ph-d-defense-yichen-tu/
LOCATION:130 Hanes Hall\, Hanes Hall\, Chapel Hill\, 27599\, United States
CATEGORIES:PhD Defense
END:VEVENT
BEGIN:VEVENT
DTSTART;TZID=America/Halifax:20190813T090000
DTEND;TZID=America/Halifax:20190813T130000
DTSTAMP:20211208T010014
CREATED:20190321T084221Z
LAST-MODIFIED:20190321T084221Z
UID:9852-1565686800-1565701200@stor.unc.edu
SUMMARY:CWE exams
DESCRIPTION:Today’s exam: 634 & 635
URL:https://stor.unc.edu/event/cwe-exams-7/
LOCATION:107 Hanes Hall\, Hanes Hall\, Chapel Hill\, NC\, 27599\, United States
END:VEVENT
BEGIN:VEVENT
DTSTART;TZID=America/Halifax:20190814T090000
DTEND;TZID=America/Halifax:20190814T130000
DTSTAMP:20211208T010014
CREATED:20190321T084254Z
LAST-MODIFIED:20190321T084254Z
UID:9853-1565773200-1565787600@stor.unc.edu
SUMMARY:CWE exams
DESCRIPTION:Today’s exams: 654 & 655 or 654 & 672
URL:https://stor.unc.edu/event/cwe-exams-8/
LOCATION:107 Hanes Hall\, Hanes Hall\, Chapel Hill\, NC\, 27599\, United States
END:VEVENT
BEGIN:VEVENT
DTSTART;TZID=America/Halifax:20190815T090000
DTEND;TZID=America/Halifax:20190815T130000
DTSTAMP:20211208T010014
CREATED:20190321T084324Z
LAST-MODIFIED:20190321T084324Z
UID:9854-1565859600-1565874000@stor.unc.edu
SUMMARY:CWE exams
DESCRIPTION:Today’s exam: 641 & 642
URL:https://stor.unc.edu/event/cwe-exams-9/
LOCATION:107 Hanes Hall\, Hanes Hall\, Chapel Hill\, NC\, 27599\, United States
END:VEVENT
BEGIN:VEVENT
DTSTART;TZID=America/Halifax:20190816T090000
DTEND;TZID=America/Halifax:20190816T130000
DTSTAMP:20211208T010014
CREATED:20190321T084355Z
LAST-MODIFIED:20190321T084355Z
UID:9855-1565946000-1565960400@stor.unc.edu
SUMMARY:CWE exams
DESCRIPTION:Today’s exam: 664 & 665
URL:https://stor.unc.edu/event/cwe-exams-10/
LOCATION:107 Hanes Hall\, Hanes Hall\, Chapel Hill\, NC\, 27599\, United States
END:VEVENT
BEGIN:VEVENT
DTSTART;TZID=America/Halifax:20190816T140000
DTEND;TZID=America/Halifax:20190816T160000
DTSTAMP:20211208T010014
CREATED:20190809T081351Z
LAST-MODIFIED:20190809T081351Z
UID:9883-1565964000-1565971200@stor.unc.edu
SUMMARY:Ph.D. defense: Zheqi Zhang
DESCRIPTION:Prioritization and Distribution of Casualties in Disaster Management \n(Under the direction of Nilay Tanik Argon and Serhan Ziya) \n \nThis dissertation focused on two different problems which typically arise in the aftermath of disasters. In the first part of the dissertation\, we study the problem of how casualties should be prioritized and distributed to different medical facilities in the aftermath of mass casualty incidents (MCIs) with the objective of maximizing the expected total number of survivors. Assuming that casualties have been triaged into two classes differentiated by their severity levels and medical needs\, the decision-maker needs to prioritize and distribute casualties using a limited number of ambulances to multiple medical facilities with different capacities. By explicitly taking into consideration the capacity and service time at each medical facility\, we formulate this sequential decision-making problem as a Markov decision process (MDP). Based on this MDP formulation\, we propose heuristic policies that prescribe decisions on prioritization and distribution of casualties. We then employ discrete-event simulations to demonstrate the benefits of using the proposed heuristics against some benchmark policies under several realistic mass casualty incident scenarios such as terrorist attacks\, major traffic accidents\, and earthquakes. \n \nIn the second part of the dissertation\, we study the resource allocation problem in urban search and rescue operations that follow natural disasters. Specifically\, we consider a scenario in which some individuals are trapped at various locations within a geographical area and there is a limited time window during which these individuals can be rescued. We model the problem as an MDP. Then\, we characterize the optimal policy under the assumption that individuals belong to only one of two locations. We propose heuristics for the general version of the problem. Finally\, the proposed heuristics are examined with a simulation.
URL:https://stor.unc.edu/event/ph-d-defense-zheqi-zhang/
LOCATION:107 Hanes Hall\, Hanes Hall\, Chapel Hill\, NC\, 27599\, United States
CATEGORIES:PhD Defense
END:VEVENT
BEGIN:VEVENT
DTSTART;TZID=America/Halifax:20190819T093000
DTEND;TZID=America/Halifax:20190819T120000
DTSTAMP:20211208T010014
CREATED:20190718T100129Z
LAST-MODIFIED:20190718T100129Z
UID:9882-1566207000-1566216000@stor.unc.edu
SUMMARY:Department Orientation
DESCRIPTION:Orientation for all incoming graduate students will be held in the 3rd floor lounge.
URL:https://stor.unc.edu/event/department-orientation/
LOCATION:Hanes Lounge\, Hanes Hall\, Chapel Hill\, NC\, 27599\, United States
END:VEVENT
BEGIN:VEVENT
DTSTART;TZID=America/Halifax:20190918T153000
DTEND;TZID=America/Halifax:20190918T163000
DTSTAMP:20211208T010014
CREATED:20190906T085653Z
LAST-MODIFIED:20190906T085653Z
UID:9887-1568820600-1568824200@stor.unc.edu
SUMMARY:Graduate Roundtable
DESCRIPTION:This event will provide the opportunity for current STOR graduate students to hear more about what is expected of them over the life of their program\, what skills they should be trying to develop\, ideas they should be thinking about\, and the kinds of habits they should develop to become successful researchers. Faculty members will also talk about other aspects of being a professor\, like writing papers\, advising students\, editing journals\, organizing conferences\, and other professional service while maintaining work life balance. Finally\, hear about how the faculty views our department\, statistics programs in general\, and even the future of the field of statistics and operations research. \nAll STOR graduate students are encouraged to attend\, but this event may be of interest to graduate students in other departments\, or undergraduates interested in going to graduate school. Students who have specific questions to ask professors should send those questions to Benjy Leinwand. \n \nRefreshments will be served at 3:00pm in the 3rd floor lounge of Hanes Hall
URL:https://stor.unc.edu/event/graduate-roundtable/
LOCATION:120 Hanes Hall\, Hanes Hall\, Chapel Hill\, NC\, 27599\, United States
END:VEVENT
BEGIN:VEVENT
DTSTART;TZID=America/Halifax:20190920T153000
DTEND;TZID=America/Halifax:20190920T163000
DTSTAMP:20211208T010014
CREATED:20190917T124249Z
LAST-MODIFIED:20190917T124249Z
UID:9890-1568993400-1568997000@stor.unc.edu
SUMMARY:Grad Student Seminar: Michael Conroy & Adam Waterbury
DESCRIPTION:Michael Conroy- UNC-Chapel Hill \nA direct Approach to Renewal Theory on Trees \nIn a variety of applications ranging from computer science to statistical physics\, a class of recursive stochastic fixed-point equations appear. Such recursions admit so-called special endogenous solutions constructed on a random weighted tree formed from i.i.d. copies of the inputs to the recursion. We are interested in the tail behavior of the endogenous solution a max-type recursion that arises in the analysis of the branching random walk and also as the limiting waiting time distribution on parallel queueing networks with synchronization. The particular form of this equation allows us to analyze the tail behavior of the solutions by extending classical change-of-measure and renewal-theoretic techniques to random trees. Our techniques offer a formulation of this tail behavior that allows for efficient simulation of tail probabilities via an importance sampling algorithm. \n \nAdam Waterbury– UNC-Chapel Hill\nWeak Limits of Quasi-Stationary Distributions \nA question of great importance in ecology is what conditions must be met in order for a population of interacting and\, possibly\, competing species to coexist with one another over long time spans. In reality if the population is finite\, then after a large enough amount of time has passed\, one or more of the species are sure to face extinction. However\, the time that it takes for extinction to occur can be quite large\, so it is natural to consider whether the population can sustain any long-term coexistence before any of the species are extinct. If the dynamics of a population are modeled by a Markov process\, then such metastability is captured in the notion of a quasi-stationary distribution. In this work we analyze the limiting behavior of quasi-stationary distributions of a family of Markov chains that model the evolution of interacting biological populations. In particular\, we show that under some large deviations assumptions\, the support of weak limit points of the quasi-stationary distributions can be described in terms of dynamical properties of the law of large numbers limit of the processes.
URL:https://stor.unc.edu/event/grad-student-seminar-michael-conroy-adam-waterbury/
LOCATION:120 Hanes Hall\, Hanes Hall\, Chapel Hill\, NC\, 27599\, United States
CATEGORIES:Graduate Seminar
END:VEVENT
BEGIN:VEVENT
DTSTART;TZID=America/Halifax:20190923T153000
DTEND;TZID=America/Halifax:20190923T163000
DTSTAMP:20211208T010014
CREATED:20190830T152333Z
LAST-MODIFIED:20190830T152333Z
UID:9885-1569252600-1569256200@stor.unc.edu
SUMMARY:Colloquium: Peter J. Mucha\, UNC-Chapel Hill
DESCRIPTION:Peter J. Mucha \nThe University of North Carolina at Chapel Hill \nDepartment of Mathematics \nCommunities in Multilayer Networks \nCommunity detection describes the organization of a network in terms of patterns of connection\, identifying tightly connected structures known as communities. A wide variety of methods for community detection have been proposed\, with a number of software packages available for performing community detection. In the past decade\, there has been increased interest in multilayer networks\, a general framework that can be used to describe networks with multiple types of relationships\, that change in time\, or that network together multiple kinds of networks. We describe various generalizations of community detection to multilayer networks\, including results about detectability limits and a new post-processing procedure to explore the parameter space of multilayer modularity\, along with pointers to using community detection in applications.
URL:https://stor.unc.edu/event/colloquium-peter-j-mucha-unc-chapel-hill/
LOCATION:120 Hanes Hall\, Hanes Hall\, Chapel Hill\, NC\, 27599\, United States
CATEGORIES:STOR Colloquium
END:VEVENT
BEGIN:VEVENT
DTSTART;TZID=America/Halifax:20190927T153000
DTEND;TZID=America/Halifax:20190927T163000
DTSTAMP:20211208T010014
CREATED:20190923T105720Z
LAST-MODIFIED:20190923T105720Z
UID:9891-1569598200-1569601800@stor.unc.edu
SUMMARY:Grad Student Seminar: Jack Prothero & Marie Dueker
DESCRIPTION:Jack Prothero – UNC-Chapel Hill\nExtracting Signal and Noise from Large Matrices \nDiscerning a low-rank signal from a large\, noisy data matrix is a classic problem in statistics and signal processing. We review a fundamental result in random matrix theory and how it applies to recent results on optimal singular value thresholding and shrinkage for signal extraction. In toy examples we find that these thresholding procedures often fail to capture as much signal as is theoretically possible. Our main contribution is a diagnostic quantile-quantile graphic for evaluating signal extraction quality. Such a diagnostic is reliant on an appropriate estimate of remaining noise after signal is extracted from a matrix. For non-square matrices this noise matrix estimation is surprisingly nontrivial; we discuss the challenges that arise and propose potential solutions to these challenges. \n \nMarie Dueker – Ruhr-Universitaet Bochum\nCommon deterministic trends in varying means model \nFor possibly nonstationary multivariate data collected over time\, common deterministic trends are linear combinations of different components that are stationary over time. The trend in mean in the underlying model is determined by a vector of deterministic functions. Then\, the largest number of linearly independent linear combinations that lead to stationarity in mean is referred to as cotrending dimension\, and their spanned space as cotrending space. The cotrending dimension and the cotrending space are related to the eigenstructure and eigenspace\, respectively\, of suitable matrices. Testing procedures for both dimension and space are presented. The talk will also discuss a possible change in the cotrending dimension over time. The results are illustrated by some simulations and data examples.
URL:https://stor.unc.edu/event/grad-student-seminar-jack-prothero-marie-dueker/
LOCATION:120 Hanes Hall\, Hanes Hall\, Chapel Hill\, NC\, 27599\, United States
CATEGORIES:Graduate Seminar
END:VEVENT
BEGIN:VEVENT
DTSTART;TZID=America/Halifax:20191004T153000
DTEND;TZID=America/Halifax:20191004T163000
DTSTAMP:20211208T010014
CREATED:20190926T135045Z
LAST-MODIFIED:20190926T135045Z
UID:9892-1570203000-1570206600@stor.unc.edu
SUMMARY:STOR Colloquium: Tong Wang\, University of Iowa
DESCRIPTION:Dr. Tong Wang \nTippie College of Business \nUniversity of Iowa \n \nHybrid Predictive Model: When an Interpretable Model Collaborates with a Black-box Model \n \nInterpretable machine learning has received increasing interest in recent years\, especially in domains where humans are involved in the decision-making process. However\, the possible loss of the task performance for gaining interpretability is often inevitable\, especially for large datasets or complicated tasks. This performance downgrade puts practitioners in a dilemma of choosing between a top-performing black-box model with no explanations and an interpretable model with unsatisfying task performance. In this work\, we propose a novel framework for building a Hybrid Predictive Model (HPM) that integrates an interpretable model with any black-box model to introduce interpretability in the decision-making process at no or low cost of the predictive accuracy. The interpretable model substitutes the black-box model on a subset of data where the black-box model is overkill or nearly overkill. We design a principled objective function that considers predictive accuracy\, model interpretability\, and model transparency\, which is the percentage of data processed by the interpretable model. This framework brings together the advantages of the high predictive performance of black-box models and the high interpretability of interpretable models. We instantiate the proposed framework with two types of models\, one using decision rules as the interpretable collaborator and one using linear models. For both models\, we develop customized training algorithms with theoretically grounded bounds to reduce computation. We test the hybrid predictive models on structured datasets and text data. In these experiments\, the interpretable models collaborate with state-of-the-art black-box models including ensemble models and neural networks. We propose to use efficient frontiers to characterize the trade-off between transparency and predictive performance. Results show that hybrid models are able to obtain transparency at no or low cost of predictive performance. \n \nRefreshments will be served at 3:00pm in the 3rd floor lounge of Hanes Hall \n
URL:https://stor.unc.edu/event/stor-colloquium-tong-wang-university-of-iowa/
LOCATION:120 Hanes Hall\, Hanes Hall\, Chapel Hill\, NC\, 27599\, United States
CATEGORIES:STOR Colloquium
END:VEVENT
BEGIN:VEVENT
DTSTART;TZID=America/New_York:20191007T153000
DTEND;TZID=America/New_York:20191007T163000
DTSTAMP:20211208T010014
CREATED:20191003T150658Z
LAST-MODIFIED:20191003T150658Z
UID:9894-1570462200-1570465800@stor.unc.edu
SUMMARY:Statistics Seminar: Wen Zhou\, Colorado State University
DESCRIPTION:Estimation and Inference of a Heteroskedasticity Model with Latent Semiparametric Factors for Panel Data Analysis \n \nWe consider estimation and inference of a flexible subject-specific heteroskedasticity model for analyzing large scale panel data\, which employs latent semiparametric factor structure to simultaneously account for the heteroskedasticity across subjects and contemporaneous correlations. Specifically\, the heteroskedasticity across subjects is modeled by the product of unobserved stationary process of factors and subject-specific covariate effect. Serving as the loading\, the covariate effect is further modeled through the additive model. We propose a two-step procedure for estimation. First\, the latent factor process and nonparametric loading are estimated via projection-based methods. The estimation of regression coefficients is further conducted through the generalized least squares type approach. Theoretical validity of the two-step procedure is carefully documented. By scrupulously examining the non-asymptotic rates for recovering the latent factor process and its loading\, we further study the properties of the estimated regression coefficients. In particular\, we establish the asymptotic normality of the proposed two-step estimate of regression coefficients. The proposed regression coefficient estimator is also shown to be asymptotically efficient. This leads to a more efficient confidence set of the regression coefficients. Using a comprehensive simulation study\, we demonstrate the finite sample performance of the proposed procedure\, and numerical results corroborate our theoretical findings. Finally\, we apply our proposed method to a data set of air quality and energy consumption collected at 129 monitoring sites in the United States in 2015. This is a collaborative work with Lvuou Zhang and Haonan Wang. \n \nRefreshments will be served at 3:00pm in the 3rd floor lounge of Hanes Hall
URL:https://stor.unc.edu/event/statistics-seminar-wen-zhou-colorado-state-university/
LOCATION:Hanes Hall\, Hanes Hall\, Chapel Hill\, NC\, 27599\, United States
END:VEVENT
BEGIN:VEVENT
DTSTART;TZID=America/New_York:20191010T161500
DTEND;TZID=America/New_York:20191010T171500
DTSTAMP:20211208T010014
CREATED:20191007T100235Z
LAST-MODIFIED:20191007T100235Z
UID:9896-1570724100-1570727700@stor.unc.edu
SUMMARY:Probability Seminar: Sayan Banerjee
DESCRIPTION:Non-parametric change point detection in growing networks \nMotivated by applications of modeling both real world and probabilistic systems such as recursive trees\, the last few years have seen an explosion in models for dynamically evolving networks. In this talk\, we consider models of growing networks which evolve via new vertices attaching to the pre-existing network according to one attachment function f till the system grows to size τ(n) < n\, when new vertices switch their behavior to a different function g till the system reaches size n. We explore the effect of this change point on the evolution and final degree distribution of the network. In particular\, we consider two cases\, the standard model where τ(n) = γn as well as the quick big bang model when τ(n) = nγ for some 0 < γ < 1. In the former case\, we obtain deterministic ‘fluid limits’ to track the degree evolution in the sup-norm metric. In the latter case\, we show that the effect of the pre-change point dynamics ‘washes out’ when the network reaches size n\, although the maximal degree feels the effect of the change. We also devise non-parametric\, consistent estimators to detect the change point. Our methods exploit and develop new techniques connecting inhomogeneous continuous time branching processes (CTBP) to the evolving networks. This is joint work with Shankar Bhamidi and Iain Carmichael.
URL:https://stor.unc.edu/event/probability-seminar-sayan-banerjee/
LOCATION:125 Hanes Hall\, Hanes Hall\, Chapel Hill\, NC\, 27599\, United States
CATEGORIES:Probability Seminar
END:VEVENT
BEGIN:VEVENT
DTSTART;TZID=America/New_York:20191011T153000
DTEND;TZID=America/New_York:20191011T163000
DTSTAMP:20211208T010014
CREATED:20191007T100040Z
LAST-MODIFIED:20191007T100040Z
UID:9895-1570807800-1570811400@stor.unc.edu
SUMMARY:Grad Student Seminar: Michael Conroy
DESCRIPTION:Asymptotic optimality in resource sharing networks \nIn this talk we consider a family of resource sharing networks that can model Internet flows when control policies are imposed to allocate resources. A fundamental problem for such systems is to construct optimal policies under admissibility constraints\, where optimality is formulated in terms of a cost function. This problem of optimizing for stochastic controls is in general intractable\, but in a heavy traffic regime (i.e. when system capacity is approximately balanced with system load)\, these stochastic control problems can be approximated by so-called Brownian control problems. Essentially what this means is that the state process of the original queue\, when rescaled\, has a limiting diffusion approximation that is easier to analyze. For both a discounted cost and an ergodic cost criterion\, the appropriate diffusion approximation gives a lower bound on the best achievable asymptotic cost under any sequence of admissible policies. Consequently\, these bounds show that the particular control policies constructed previously in Budhiraja and Johnson (2015) are in fact asymptotically optimal.
URL:https://stor.unc.edu/event/grad-student-seminar-michael-conroy/
LOCATION:120 Hanes Hall\, Hanes Hall\, Chapel Hill\, NC\, 27599\, United States
CATEGORIES:Graduate Seminar
END:VEVENT
BEGIN:VEVENT
DTSTART;TZID=America/Halifax:20191014T153000
DTEND;TZID=America/Halifax:20191014T163000
DTSTAMP:20211208T010014
CREATED:20190911T080737Z
LAST-MODIFIED:20190911T080737Z
UID:9888-1571067000-1571070600@stor.unc.edu
SUMMARY:STOR Colloquium: Heping Zhang\, Yale
DESCRIPTION:Back to the Basics: Residuals and Diagnostics for Generalized Linear Models\nHeping Zhang \nSusan Dwight Bliss Professor of Biostatistics \nYale University School of Public Health \nOrdinal outcomes are common in scientific research and everyday practice\, and we often rely on regression models to make inference. A long-standing problem with such regression analyses is the lack of effective diagnostic tools for validating model assumptions. The difficulty arises from the fact that an ordinal variable has discrete values that are labeled with\, but not\, numerical values. The values merely represent ordered categories. In this paper\, we propose a surrogate approach to defining residuals for an ordinal outcome Y. The idea is to define a continuous variable S as a “surrogate” of Y and then obtain residuals based on S. For the general class of cumulative link regression models\, we study the residual’s theoretical and graphical properties. We show that the residual has null properties similar to those of the common residuals for continuous outcomes. Our numerical studies demonstrate that the residual has power to detect misspecification with respect to 1) mean structures; 2) link functions; 3) heteroscedasticity; 4) proportionality; and 5) mixed populations. The proposed residual also enables us to develop numeric measures for goodness-of-fit using classical distance notions. Our results suggest that compared to a previously defined residual\, our residual can reveal deeper insights into model diagnostics. We stress that this work focuses on residual analysis\, rather than hypothesis testing. The latter has limited utility as it only provides a single p-value\, whereas our residual can reveal what components of the model are misspecified and advise how to make improvements. \nThis is a joint work with Dungang Liu\, University of Cincinnati Lindner College of Business. \nThe entire article can be viewed here.
URL:https://stor.unc.edu/event/stor-colloquium-heping-zhang-yale/
LOCATION:120 Hanes Hall\, Hanes Hall\, Chapel Hill\, NC\, 27599\, United States
CATEGORIES:STOR Colloquium
END:VEVENT
BEGIN:VEVENT
DTSTART;TZID=America/Halifax:20191021T153000
DTEND;TZID=America/Halifax:20191021T163000
DTSTAMP:20211208T010014
CREATED:20190926T135335Z
LAST-MODIFIED:20190926T135335Z
UID:9893-1571671800-1571675400@stor.unc.edu
SUMMARY:STOR Colloquium: Shujie Ma\, UC Riverside
DESCRIPTION:Shujie Ma \nUniversity of California\, Riverside \n \nHow many communities are there in a network? \n \nAdvances in modern technology have facilitated the collection of network data which emerge in many fields including biology\, bioinformatics\, physics\, economics\, sociology and so forth. Network data often have natural communities which are groups of interacting objects (i.e.\, nodes); pairs of nodes in the same group tend to interact more than pairs belonging to different groups. Community detection then becomes a very important task\, allowing us to identify and understand the structure of a network. Thus\, the development of methods for community detection has attracted much attention in the past decade\, and as a result\, different efficient approaches have been proposed in literature. \n \nA fundamental limitation of most existing methods is that they divide networks into a fixed number of communities\, i.e.\, the number of communities is known and given in advance. However\, in practice\, such prior information is typically unavailable. Determining the number of communities is a challenging yet important task\, as the following community detection procedure relies upon it. In this talk\, I will introduce a convenient and effective solution to this problem under the degree-corrected stochastic block models (DC-SBM). The proposed method takes advantages of spectral clustering\, likelihood principle and binary segmentation. Determining the number of communities is essentially a model selection problem\, and we therefore establish the selection consistency of our proposed procedure under a mild condition on the average degree. We demonstrate the approach on different networks. At the end of my talk\, I will briefly talk about our other on-going and future research projects in this line of work. \n \n \n \nRefreshments will be served at 3:00pm in the 3rd floor lounge of Hanes Hall
URL:https://stor.unc.edu/event/stor-colloquium-shujie-ma-uc-riverside/
LOCATION:120 Hanes Hall\, Hanes Hall\, Chapel Hill\, NC\, 27599\, United States
CATEGORIES:STOR Colloquium
END:VEVENT
BEGIN:VEVENT
DTSTART;TZID=America/New_York:20191023T190000
DTEND;TZID=America/New_York:20191023T200000
DTSTAMP:20211208T010014
CREATED:20191016T155934Z
LAST-MODIFIED:20191016T155934Z
UID:9898-1571857200-1571860800@stor.unc.edu
SUMMARY:Undergrad Advising
DESCRIPTION:Advising night for any STAN majors or minors who would like the opportunity to speak to a departmental advisor. Spring registration begins November 4th!
URL:https://stor.unc.edu/event/undergrad-advising/
LOCATION:120 Hanes Hall\, Hanes Hall\, Chapel Hill\, NC\, 27599\, United States
END:VEVENT
BEGIN:VEVENT
DTSTART;TZID=America/New_York:20191024T161500
DTEND;TZID=America/New_York:20191024T171500
DTSTAMP:20211208T010014
CREATED:20191015T083344Z
LAST-MODIFIED:20191015T083344Z
UID:9897-1571933700-1571937300@stor.unc.edu
SUMMARY:Optimization Seminar: Gabor Pataki
DESCRIPTION:Proving infeasibility in semidefinite programming: how elementary row operations help \n Gabor Pataki\, University of North Carolina at Chapel Hill \n Semidefinite programs (SDPs) — optimization problems with linear constraints and semidefinite matrix variables — are some of the most useful\, versatile\, and popular optimization problems of the last three decades. \nA fundamental\, and difficult issue in SDP is proving infeasibility. SDP has a kind of Farkas’ lemma\, i.e.\, a certain alternative system\, but this often fails to prove infeasibility. \nI will show that we can always prove infeasibility using a surprisingly simple tool: we can transform SDPs to a normal form that makes the infeasibility trivial to recognize. The transformation is very simple\, as it mostly uses elementary row operations coming from Gaussian elimination\, and it builds on the idea of facial reduction. Thus our normal form of SDPs is analogous to the row echelon form of a linear system of equations. \nThe normal form has theoretical uses: for example\, it gives a simple proof that SDP feasibility is in co-NP in the real number model of computing. The normal form also has computational uses: in many cases it helps to recognize infeasibility in practice. \nThe talk relies only on knowledge of elementary linear algebra. Most of this work is joint with Minghui Liu. Some other parts are joint work with Yuzixuan Zhu and Quoc Tran-Dinh.
URL:https://stor.unc.edu/event/optimization-seminar-gabor-pataki/
CATEGORIES:Optimization Seminar
END:VEVENT
BEGIN:VEVENT
DTSTART;TZID=America/New_York:20191025T153000
DTEND;TZID=America/New_York:20191025T163000
DTSTAMP:20211208T010014
CREATED:20191025T105305Z
LAST-MODIFIED:20191025T105305Z
UID:9899-1572017400-1572021000@stor.unc.edu
SUMMARY:Grad Student Seminar: Miheer Dewaskar
DESCRIPTION:Miheer Dewaskar \nUNC-Chapel Hill \n \nAsymptotic analysis of the power of choice phenomenon \nSuppose that n balls are to be sequentially placed into n bins with the objective of keeping the maximum load of the bins small. In absence of a central dispatcher\, and in order to minimize the communication overhead\, each incoming ball chooses d bins uniformly at random and goes into the bin with the smallest load among its d choices. The maximum bin load for d = 2 (or greater) is substantially smaller than that for d=1 : O(log log n) vs O(log n); this phenomenon is called the power of choice. The phenomenon is quite robust and has applications to large scale load balancing\, hashing\, collision protocols\, etc. In this talk we will analyze the mean behavior of the balls and bins problem described above\, and use concentration inequalities to show that\, with high probability\, the maximum load is (log log n)/log d +/- 1 when d > 1.
URL:https://stor.unc.edu/event/grad-student-seminar-miheer-dewaskar/
LOCATION:120 Hanes Hall\, Hanes Hall\, Chapel Hill\, NC\, 27599\, United States
CATEGORIES:Graduate Seminar
END:VEVENT
BEGIN:VEVENT
DTSTART;TZID=America/New_York:20191106T153000
DTEND;TZID=America/New_York:20191106T163000
DTSTAMP:20211208T010014
CREATED:20190903T144239Z
LAST-MODIFIED:20190903T144239Z
UID:9886-1573054200-1573057800@stor.unc.edu
SUMMARY:STOR Colloquium: Jianqing Fan\, Princeton
DESCRIPTION:Jianqing Fan \nPrinceton University \n \nStatistical Inference on Membership Profiles in Large Networks \n \nNetwork data is prevalent in many contemporary big data applications in which a common interest is to unveil important latent links between different pairs of nodes. The nodes can be broadly defined such as individuals\, economic entities\, documents\, or medical disorders in social\, economic\, text\, or health networks. Yet a simple question of how to precisely quantify the statistical uncertainty associated with the identification of latent links still remains largely unexplored. In this talk\, we suggest the method of statistical inference on membership profiles in large networks (SIMPLE) in the setting of degree-corrected mixed membership model\, where the null hypothesis assumes that the pair of nodes share the same profile of community memberships. In the simpler case of no degree heterogeneity\, the model reduces to the mixed membership model and an alternative more robust test is proposed. Under some mild regularity conditions\, we establish the exact limiting distributions of the two forms of SIMPLE test statistics under the null hypothesis and their asymptotic properties under the alternative hypothesis. Both forms of SIMPLE tests are pivotal and have asymptotic size at the desired level and asymptotic power one. The advantages and practical utility of our new method in terms of both size and power are demonstrated through several simulation examples and real network applications. \n(Joint work with Yingying Fan\, Xiao Han\, and Jinchi Lv) \n \nThe talk is based on the following paper on arxiv.org \nFan\, J.\, Fan\, Y.\, Han\, X. and Lv\, J. (2019). SIMPLE: Statistical Inference \non Membership Profiles in Large Networks
URL:https://stor.unc.edu/event/stor-colloquium-jainqing-fan-princeton/
LOCATION:120 Hanes Hall\, Hanes Hall\, Chapel Hill\, NC\, 27599\, United States
CATEGORIES:STOR Colloquium
END:VEVENT
BEGIN:VEVENT
DTSTART;TZID=America/New_York:20191106T191500
DTEND;TZID=America/New_York:20191106T201500
DTSTAMP:20211208T010014
CREATED:20191104T124542Z
LAST-MODIFIED:20191104T124542Z
UID:9901-1573067700-1573071300@stor.unc.edu
SUMMARY:CASO Career Panel
DESCRIPTION:Carolina Actuarial Student Organization (CASO) is inviting two actuaries to talk about their work in different fields. Dev Patel\, FCAS\, is a former actuary working in population health programs. Fred Peterson\, FSA\, will talk about pensions and retirement products. Please feel free to swing in at Hanes 125 at 7:15pm on Wednesday (Nov.6th) to find out more! Free pizza and refreshment served.
URL:https://stor.unc.edu/event/caso-career-panel/
LOCATION:125 Hanes Hall\, Hanes Hall\, Chapel Hill\, NC\, 27599\, United States
END:VEVENT
BEGIN:VEVENT
DTSTART;TZID=America/New_York:20191107T161500
DTEND;TZID=America/New_York:20191107T171500
DTSTAMP:20211208T010014
CREATED:20191104T124343Z
LAST-MODIFIED:20191104T124343Z
UID:9900-1573143300-1573146900@stor.unc.edu
SUMMARY:Probability Seminar: Souvik Dhara\, MIT
DESCRIPTION:Souvik Dhara\nMIT \nA new universality class for critical percolation on networks with heavy-tailed degrees \nThe talk concerns critical behavior of percolation on finite random networks with heavy-tailed degree distribution. In a seminal paper\, Aldous (1997) identified the scaling limit for the component sizes in the critical window of phase transition for the Erdős-Rényi random graph. Subsequently\, there has been a surge in the literature identifying two universality classes for the critical behavior depending on whether the asymptotic degree distribution has a finite or infinite third moment. \nIn this talk\, we will present a completely new universality class that arises in the context of degrees having infinite second moment. Specifically\, the scaling limit of the rescaled component sizes is different from the general description of multiplicative coalescent given by Aldous and Limic (1998). Moreover\, the study of critical behavior in this regime exhibits several surprising features that have never been observed in any other universality classes so far. \nThis is based on joint works with Shankar Bhamidi\, Remco van der Hofstad\, Johan van Leeuwaarden. \n \nRefreshments will be served at 3:45 in the 3rd floor lounge of Hanes Hall
URL:https://stor.unc.edu/event/probability-seminar-souvik-dhara-mit/
LOCATION:125 Hanes Hall\, Hanes Hall\, Chapel Hill\, NC\, 27599\, United States
CATEGORIES:Probability Seminar
END:VEVENT
BEGIN:VEVENT
DTSTART;TZID=America/New_York:20191120T171500
DTEND;TZID=America/New_York:20191120T181500
DTSTAMP:20211208T010014
CREATED:20190911T080924Z
LAST-MODIFIED:20190911T080924Z
UID:9889-1574270100-1574273700@stor.unc.edu
SUMMARY:Statistics Seminar: Mike Baiocchi\, Stanford
DESCRIPTION:Mike Baiocchi \nStanford University \n \nWe need better tools to solve these problems – several statistical wins in the fight against sexual assault \n \nOur team’s research on preventing sexual assault in Kenya has led to the development of several new statistical techniques to overcome challenges inherent in contexts where behaviors and specifically behavior-change are of great importance. The techniques include: (i) a method for designing randomized trials when you anticipate “spillover” or “contamination” between the intervention groups\, (ii) a method for using open-response/free-text and getting “confidence sets” and “p-values” that rigorously assess the causal differences between the two groups\, and (iii) a geo-spatial\, mixed methods approach to understanding and communicating the burden of gender based violence. The goal of this talk is to present cool new methods that are useful for you and your research. Lots of intuition and pictures will be used. \nBiosketch: Michael Baiocchi\, PhD\, is an Assistant Professor in the Stanford Prevention Research Center. He is an interventional-statistician\, creating interventions and the means for analyzing them. He specializes in creating simple\, easy to understand statistical methodologies for getting reliable results out of messy data and messy situations. His research is in nonparametric estimation and design-based inference. He was the inaugural Stein Fellow in the department of Statistics at Stanford University. He is the principal investigator on a large (enrollment: 5\,000+ students\, 100+ schools) randomized study of a sexual assault prevention intervention in the informal settlements around Nairobi\, Kenya. He is currently launching a five-university study of a sexual assault resistance program in the United States.
URL:https://stor.unc.edu/event/probability-seminar-mike-baiocchi-stanford/
LOCATION:120 Hanes Hall\, Hanes Hall\, Chapel Hill\, NC\, 27599\, United States
CATEGORIES:Statistics Seminar
END:VEVENT
BEGIN:VEVENT
DTSTART;TZID=America/New_York:20191122T153000
DTEND;TZID=America/New_York:20191122T163000
DTSTAMP:20211208T010014
CREATED:20191120T091051Z
LAST-MODIFIED:20191120T091051Z
UID:9902-1574436600-1574440200@stor.unc.edu
SUMMARY:Grad Student Seminar: Carson Mosso\, Jonghwan Yoo
DESCRIPTION:Carson Mosso – Latent Association Mining in Binary Data \n \nWe will introduce a new data mining method for binary valued data\, called Latent Association Mining in Binary Data. The origin of this problem is in market basket analysis\, where binary valued data is common\, and typically falls under the branch of data mining called Association Rule Mining. However\, the problem can be generalized by mining for correlation between variables in various types of datasets\, e.g.\, text or gene expression data. First\, we will introduce a latent variable model and define a new statistic called latent association. This statistic is similar to correlation\, but better suited to our latent variable model. Then we will define the association structure that we are interested in mining and an iterative hypothesis testing algorithm to find this association structure. This talk will discuss both statistical theory and real data applications. Moreover\, we will spend time introducing Association Rule Mining\, to put the problem in its proper context\, and discuss two related latent variable methods\, Latent Dirichlet Allocation and Nonnegative Matrix Factorization. \n \nJonghwan Yoo – Integrative Data Analysis on H&E Images and Methylation Data \n \nDue to advances in technology\, various types of data for a common set of subjects or samples have become available. For instance\, genetic\, genomic\, epigentic\, neurocognitive\, clinical and image data can be collected for each subject. Each of these data provides shared or partly independent information represented in different ways. Integrating information from various data sources\, therefore\, is essential to gain a broad and deeper understanding of subjects. One effective approach is to use Angle-Based Joint and Individual Variation Explained (AJIVE). We apply AJIVE on H&E images and methylation data of a skin cancer dataset. A main challenge is the enormous size of the H&E images and this is tackled by a deep learning technique called automated feature extraction. \n
URL:https://stor.unc.edu/event/grad-student-seminar-carson-mosso-jonghwan-yoo/
LOCATION:120 Hanes Hall\, Hanes Hall\, Chapel Hill\, NC\, 27599\, United States
CATEGORIES:Graduate Seminar
END:VEVENT
BEGIN:VEVENT
DTSTART;TZID=America/New_York:20191205T161500
DTEND;TZID=America/New_York:20191205T171500
DTSTAMP:20211208T010014
CREATED:20191203T080620Z
LAST-MODIFIED:20191203T080620Z
UID:9903-1575562500-1575566100@stor.unc.edu
SUMMARY:Probability Seminar: Zoe Huang\, Duke
DESCRIPTION:Zoe Huang\nDuke \nThe contact process on Galton-Watson trees \n \nAbstract: The contact process describes an epidemic model where each \ninfected individual recovers at rate 1 and infects its healthy neighbors \nat rate $\lambda$. We show that for the contact process on Galton-Watson \ntrees\, when the offspring distribution (i) is subexponential the \ncritical value for local survival $\lambda_2=0$ and (ii) when it is \nGeometric($p$) we have $\lambda_2 \le C_p$\, where the $C_p$ are much \nsmaller than previous estimates. This is based on an improved (and in a \nsense sharp) understanding of the survival time of the contact process \non star graphs. Recently it is proved by Bhamidi\, Nam\, Nguyen and Sly \n(2019) that when the offspring distribution of the Galton-Watson tree \nhas exponential tail\, the first critical value $\lambda_1$ of the \ncontact process is strictly positive. We prove that if the contact \nprocess survives then the number of infected sites grows exponentially \nfast. As a consequence we show that the contact process dies out at the \ncritical value $\lambda_1$ and does not survive strongly at $\lambda_2$. \nBased on joint work with Rick Durrett. \n \n \n \nRefreshments will be served at 3:45 in the 3rd floor lounge of Hanes Hall \n
URL:https://stor.unc.edu/event/probability-seminar-zoe-huang-duke/
CATEGORIES:Probability Seminar
END:VEVENT
BEGIN:VEVENT
DTSTART;TZID=America/New_York:20200108T153000
DTEND;TZID=America/New_York:20200108T163000
DTSTAMP:20211208T010014
CREATED:20200103T144647Z
LAST-MODIFIED:20200103T144647Z
UID:9904-1578497400-1578501000@stor.unc.edu
SUMMARY:STOR Colloquium: Anna Little\, Michigan State University
DESCRIPTION:Robust Statistical Procedures for Noisy\, \nHigh-dimensional Data \n \nThis talk addresses two topics related to robust statistical procedures for analyzing noisy\, high-dimensional data: (I) path-based spectral clustering and (II) robust multi-reference alignment. Both methods must overcome a large ambient dimension and lots of noise to extract the relevant low dimensional data structure in a computationally efficient way. In (I)\, the goal is to partition the data into meaningful groups\, and this is achieved by a novel approach which combines a data driven metric with graph-based clustering. Using a data driven metric allows for strong theoretical guarantees and fast algorithms when clusters concentrate around low-dimensional sets. In (II)\, the goal is to recover a hidden signal from many noisy observations of the hidden signal\, where each noisy observation includes a random translation\, a random dilation\, and high additive noise. A wavelet based approach is used to apply a data-driven\, nonlinear unbiasing procedure\, so that the estimate of the hidden signal is robust to high frequency perturbations. \n \nRefreshments will be served at 3:00pm in the 3rd floor lounge of Hanes Hall \n
URL:https://stor.unc.edu/event/stor-colloquium-anna-little-michigan-state-university/
LOCATION:120 Hanes Hall\, Hanes Hall\, Chapel Hill\, NC\, 27599\, United States
CATEGORIES:STOR Colloquium
END:VEVENT
END:VCALENDAR