Spring 2021 Seminars
The 2020-2021 Seminar Series will be offered virtually using Zoom instead of in person. Registration will be required to receive the zoom link for the event. Please see the registration link associated with each talk to sign up. Links are specific to individual talks, so you will need to register for every talk you are interested in attending. Please email Kisa Kowal at k-kowal@northwestern.edu if you have questions.
Seminar Series talks are free and open to faculty, graduate students, and advanced undergraduate students.
Detection and Estimation Problems in Combinatorial Statistics
Wednesday, March 31, 2021
Time: 11:00 a.m. - registration required
Speaker: Miklós Z. Rácz, Assistant Professor ORFE department, Affiliated Faculty Member at the Center for Statistics and Machine Learning (CSML), Princeton University
Abstract: From networks to genomics, large amounts of data are abundant and play critical roles in helping us understand complex systems. In many such settings, these data take the form of large discrete structures with important combinatorial properties. The interplay between structure and randomness in these systems presents unique statistical challenges. In this talk I will highlight these via fundamental detection and estimation problems in two settings.
First, I will discuss various problems which probe the underlying structure of networks. These include uncovering geometric structure, as well as understanding correlation in growing networks, where I will highlight connections to the influence of the seed graph in the underlying growth model.
I will also discuss trace reconstruction, a fundamental problem where the goal is to reconstruct a sequence from noisy copies. I will explain our applied work on this problem in the context of DNA data storage, an exciting emerging technology, as well as our theoretical work including approximate versions and generalizations to trees.
Assumption-Lean Analysis of Cluster Randomized Trials in Infectious Diseases
Wednesday, April 14, 2021
Time: 11:00 a.m. - registration required
Speaker: Hyunseung Kang, Assistant Professor, Department of Statistics, University of Wisconsin-Madison
Abstract: Cluster randomized trials (CRTs) are a popular design to study the effect of interventions in infectious disease settings. However, standard analysis of CRTs primarily relies on strong parametric methods, usually a Normal mixed effect models to account for the clustering structure, and focus on the overall intent-to-treat (ITT) effect to evaluate effectiveness. The paper presents two methods to analyze two types of effects in CRTs, the overall and heterogeneous ITT effects and the spillover effect among never-takers who cannot or refuse to take the intervention. For the ITT effects, we make a modest extension of an existing method where we do not impose parametric models or asymptotic restrictions on cluster size. For the spillover effect among never-takers, we propose a new bound-based method that uses pre-treatment covariates, classification algorithms, and a linear program to obtain sharp bounds. A key feature of our method is that the bounds can become dramatically narrower as the classification algorithm improves and the method may also be useful for studies of partial identification with instrumental variables. We conclude by reanalyzing a CRT studying the effect of face masks and hand sanitizers on transmission of 2008 interpandemic influenza in Hong Kong. This is joint work with Chan Park (UW-Madison).
NEW TOPIC: Causal Inference about Gun Violence Prevention
Wednesday, April 21, 2021
Time: 11:00 a.m. - registration required
Speaker: Dylan Small, Class of 1965 Wharton Professor of Statistics, Department of Statistics, The Wharton School, University of Pennsylvania
Abstract: I will discuss two attempts at making causal inferences about gun violence prevention policies. The first policy is removing guns from intimate partner violence abusers and the second policy is background checks for private gun sales. I will also briefly describe some open areas of research about gun violence prevention that statisticians might be able to contribute to.
A Flexible Regression Model for Dispersed Count Data
Wednesday, April 28, 2021
Time: 11:00 a.m. - registration required
Speaker: Kimberly F. Sellers, Professor, Department of Mathematics and Statistics, Georgetown University
Abstract: While Poisson regression serves as a standard tool for modeling the association between a count response variable and explanatory variables, its underlying equi-dispersion assumption and its implications are well documented. The Conway-Maxwell-Poisson (COM-Poisson) distribution is a flexible count data alternative that allows for data over- or under-dispersion, thus the COM-Poisson regression can flexibly model associations involving a discrete count response variable and covariates. This talk introduces the resulting regression along with its zero-inflated analog, and the associated COMPoissonReg package in R which has become a popular resource for statistical computing.
Statistical Multiplicity Adjustment in Clinical Trials
Wednesday, May 12, 2021
Time: 11:00 a.m. - registration required.
Speaker: Jiangtao Gou, Assistant Professor, Mathematics and Statistics, Villanova University
Abstract: Statistical multiplicity contributes to an increased frequency of type I errors. A good trial design of multiple testing procedures provides a proper control of the type I error rate due to a multiplicity of tests from difference sources and reduces the type II error rate at the same time. This presentation provides a concise overview of the development of multiple testing procedures, discusses the validity of several common perceptions, and introduces several recent advances in multiple testing procedures.
A statistical method for estimating (specific) causation in the law
Wednesday, May 19, 2021
Time: 11:00 a.m. - registration required
Speaker: Maria Cuellar, Assistant Professor, Criminology Department, University of Pennsylvania
Abstract: Researchers often need to determine whether a specific exposure, or something else, caused an individual's outcome. To answer questions of causality in which the exposure and outcome have already been observed, researchers have suggested estimating the probability of causation (PC). PC is especially important in court, for example in class action lawsuits, and in public and health policy, for example in determining who has benefited most from a program. However, the current estimation methods for PC make strong parametric assumptions, or are inefficient and do not easily yield inferential tools. In this talk, I will describe an influence-function-based nonparametric estimator for a projection of PC, which allows for simple interpretation and valid inference by making only weak structural assumptions. I compare my proposed estimator to the current plug-in methods, both parametric and nonparametric, by simulation. Finally, I present an application of the proposed estimator by using data from a randomized controlled trial in Western Kenya.
Perfect is the enemy of good: new shrinkage estimators for genomics
Wednesday, May 26, 2021
Time: 11:00 a.m. - registration required
Speaker: Sihai Dave Zhao, Assistant Professor, Department of Statistics, University of Illinois at Urbana-Champaign
Abstract: Simultaneous estimation problems have a long history in statistics and have become especially common and important in genomics research: modern technologies can simultaneously assay tens of thousands to even millions of genomic features that can each introduce an unknown parameter of interest. These applications reveal some conceptual and methodological gaps in the standard empirical Bayes approach to simultaneous estimation. This talk reviews some standard approaches, illustrates some difficulties, introduces an alternative approach based on regression modeling, and illustrates some new estimators that can be applied to gene expression denoising, coexpression network reconstruction, and large-scale gene expression imputation.
Back to top