Statistics and Data Science Seminars

COSAM Departments Mathematics & Statistics Research Seminars Statistics and Data Science



Upcoming Statistics and Data Science Seminars

More Events...

Past Statistics and Data Science Seminars
DMS Statistics and Data Science (SDS) Seminar
Dec 03, 2025 01:00 PM
ZOOM


madrid-padilla
Speaker: Carlos-Misael Madrid-Padilla (Washington University in St. Louis)
Title: Temporal Spatial Model via Trend Filtering
 
 
Abstract: This talk explores the estimation of nonparametric regression functions in the presence of temporal-spatial dependencies. In such a context, trend filtering, a nonparametric estimator, is studied. To the best of our knowledge, this estimator has not previously been examined in a similar context. In the univariate case, the signals considered are assumed to have a k-th weak derivative with bounded total variation, allowing for a general degree of smoothness. In the multivariate setting, we study a variant of the K-nearest neighbor fused lasso estimator. For this case, the function is required to have bounded variation and satisfy a property that extends a piecewise Lipschitz continuity criterion, or the function is assumed to be piecewise Lipschitz. To enable efficient computation, we developed an ADMM algorithm. By aligning with lower bounds, the minimax optimality of the univariate and multivariate estimators is shown. A unique phase transition phenomenon, previously unprecedented in trend filtering studies, emerges through the analysis. Both simulation studies and real data applications underscore the superior performance of the method when compared with established techniques in the existing literature.
 
Bio:
Carlos-Misael Madrid-Padilla is a tenure-track Assistant Professor in the Department of Statistics and Data Science at Washington University in St. Louis. He earned a Ph.D. in Mathematics at the Department of Mathematics at the University of Notre Dame under the supervision of Dr. Daren Wang. During the first two years of his Ph.D., he received a master's degree in mathematics under the supervision of Dr. Alex Himonas.  His research interests include high-dimensional statistics, nonparametric statistics, change point detection, causal inference, Bayes methodology, etc. 
 

Host: Haotian Xu

 

SDS seminar’s website: https://auburn.edu/cosam/datascienceseminar/index.htm

DMS Statistics and Data Science (SDS) Seminar
Nov 19, 2025 01:00 PM
ZOOM


zhang
 
Speaker: Ruizhi Zhang (University of Georgia)
 
Title: Robust Sequential Change Detection: The Approach Based on Breakdown Points and Influence Functions
Abstract: Sequential change-point detection has many important applications in industrial quality control, signal detection, and clinical trials. However, many classical procedures may fail when the observed data are contaminated by outliers, even if the percentage of outliers is very small. In this paper, we focus on the problem of robust sequential change-point detection in the presence of a small proportion of random outliers. We first study the statistical detection properties of a general family of detection procedures under Huber’s gross error model. Moreover, we incorporate ideas of the breakdown point and the influence function from the classical offline robust statistics literature and propose their new definitions to quantify the robustness of general sequential change-point detection procedures. Then, we derive the breakdown points and influence functions of our proposed family of detection procedures, which provide a quantitative analysis of the robustness of these procedures. Moreover, we find the optimal robust bounded-influence procedure in that general family that has the smallest detection delay subject to the constraints on the false alarm rate influence function. It turns out the optimal procedure is based on the truncation of the scaled likelihood ratio statistic and has a simple form. Finally, we demonstrate the robustness and the detection efficiency of the optimal robust bounded-influence procedure through extensive simulations and compute numerical approximations of breakdown points and influence functions of some procedures to have a quantitative understanding of the robustness of different procedures.
 
Bio:
Ruizhi Zhang is an Associate Professor in the Department of Statistics at the University of Georgia, Athens. Before that, he was an assistant professor in the Department of Statistics at the University of Nebraska-Lincoln. He received his B.S. degree in Mathematics from Hua Loo-Keng Talent Program in Mathematics at the University of Science and Technology of China (USTC) in 2014, graduated with honors.  He received his P.h.D degree in Industrial Engineering from the School of Industrial and Systems Engineering at Georgia Institute of Technology in 2019, and he was co-advised by Prof. Yajun Mei and Prof. Jianjun Shi.  His research interests include change-point detection, sequential analysis, robust statistics, high-dimensional statistical inference, etc. 
 
Host: Haotian Xu
DMS Statistics and Data Science Seminar
Nov 12, 2025 01:00 PM
358 Parker Hall


ahnnguyen
 
Speaker: Dr. Anh Nguyen (Computer Science Department, Auburn University)
 
Title: How to make Vision Language Models see and explain themselves
 
 
Abstract: Large Language Models, or LLMs, with their massive world knowledge learned from text, have completely changed the game. They’ve introduced a new era: vision-language models (VLMs). In these models, images and text live in the same representation space, and instead of predicting from a fixed set of labels, they draw predictions from an open vocabulary. In this talk, I will walk you through three challenges of integrating vision capabilities into LLMs: 
      First, LLMs are strongly biased, for example, they might (over)prefer the number 7 or certain names like Biden, and that bias comes straight from their training data.
      Second, a language bias is usually seen as a blessing that helps models generalize beyond training data but also a curse in vision tasks that demand careful, detailed image analysis. 
      And third, it turns out that VLMs do not have very good "eyesight" when tested on a test similar to the eye exams for humans. Because of this, VLMs can sometimes behave in ways we don’t expect, which calls for an interface that allows humans to understand the thought process of VLMs. However, there is not yet a natural way to explain VLM decisions on an image like chain of thoughts in text.
 
I’ll share my proposal for general Explainable Bottleneck, and our implementation of Part-Based Explainable and Editable Bottleneck (PEEB) networks. In fine-grained image classification, PEEB does not only explain its predictions by describing each visual part of an object, but also lets users reprogram the classifier’s logic using natural language---right at test time.
 
DMS Statistics and Data Science Seminar
Oct 29, 2025 01:00 PM
ONLINE


maweidong

Speaker: Weidong Ma (Univ. of Pennsylvania, Perelman School of Medicine, Biostatistics and Epidemiology)

Title: A Novel Framework for Addressing Disease Under-Diagnosis Using EHR Data.

 
Abstract: Effective treatment of medical conditions begins with an accurate diagnosis. However, many conditions are often underdiagnosed, either being overlooked or diagnosed after significant delays. Electronic Health Records (EHRs) contain extensive patient health information, offering an opportunity to probabilistically identify underdiagnosed individuals. The rationale is that both diagnosed and underdiagnosed patients may display similar health profiles in EHR data, distinguishing them from condition-free patients. Thus, EHR data can be leveraged to develop models that assess an individual’s risk of having a condition. To date, this opportunity has largely remained unexploited, partly due to the lack of suitable statistical methods. The key challenge is the positive-unlabeled EHR data structure, which consists of data for diagnosed ("positive") patients and the remaining ("unlabeled") that include underdiagnosed patients and many condition-free patients. Therefore, data for patients who are unambiguously condition-free, essential for developing risk assessment models, is unavailable. To overcome this challenge, we propose ascertaining condition statuses for a small subset of unlabeled patients. We develop a novel statistical method for building accurate models using this supplemented EHR data to estimate the probability that a patient has the condition of interest. Building on the developed risk prediction model, we further study the potential factors that may contribute to under-diagnosis. Numerical simulation studies and real data applications are conducted to assess the performance of the proposed methods.
 
 
Host: Huan He
 
SDS seminar’s website: https://auburn.edu/cosam/datascienceseminar/index.htm
 
DMS Statistics and Data Science Seminar
Oct 22, 2025 01:00 PM
ONLINE


 

Dmitrii

Speaker: Dmitrii Ostrovskii (Georgia Tech — Math and ISyE)

Title: Near-Optimal and Tractable Estimation under Shift-Invariance

 

Abstract: How hard is it to estimate a discrete-time signal \((x_1, \dots, x_n) \in \mathbb{C}^n\) satisfying an unknown linear recurrence relation of order s and observed in i.i.d. complex Gaussian noise? The class of all such signals is parametric but extremely rich: It contains all exponential polynomials over \(\mathbb{C}\) with total degree s, including harmonic oscillations with s arbitrary frequencies. Geometrically, this class corresponds to the projection onto \(\mathbb{C}^n\) of the union of all shift-invariant subspaces of \(\mathbb{C}^{Z}\) of dimension s. We show that the statistical complexity of this class, as measured by the squared minimax radius of the (1−δ)-confidence ℓ2-ball, is nearly the same as for the class of s-sparse signals, namely \(O(slog(en)+log(δ−1))⋅log2(es)⋅log(en/s)\). Moreover, the corresponding near-minimax estimator is tractable, and it can be used to build a test statistic with a near-minimax detection threshold in the associated detection problem. These statistical results rest upon an approximation-theoretic one: We show that finite-dimensional shift-invariant subspaces admit compactly supported reproducing kernels whose Fourier spectra have nearly the smallest possible ℓp-norms, for all \(p∈[1,+∞]\) at once. 

 

Host: Haotian Xu

SDS seminar’s website: https://auburn.edu/cosam/datascienceseminar/index.htm

DMS Statistics and Data Science Seminar
Oct 08, 2025 01:00 PM
358 Parker Hall


bo li

Speaker: Prof. Bo Li (Department of Statistics and Data Science, Washington University in St. Louis)

Title: Spatially Varying Changepoint Detection with Application to Mapping the Impact of the Mount Pinatubo Eruption

 

Abstract: Significant events such as volcanic eruptions can exert global and long-lasting impacts on climate. These impacts, however, are not uniform across space and time. Motivated by the need to understand how the 1991 Mt. Pinatubo eruption influenced global and regional climate, we propose a Bayesian framework to simultaneously detect and estimate spatially varying temporal changepoints. Our approach accounts for the diffusive nature of volcanic effects and leverages spatial correlation. We then extend the changepoint detection problem to large-scale spherical spatiotemporal data and develop a scalable method for global applications. The framework enables Gibbs sampling for changepoints within MCMC, offering greater computational efficiency than the Metropolis–Hastings algorithm. To address the high dimensionality of global data, we incorporate spherical harmonic transformations, which further substantially reduce computational burden while preserving accuracy. We demonstrate the effectiveness of our method using both simulated datasets and real data on stratospheric aerosol optical depth and surface temperature to detect and estimate changepoints associated with the Mt. Pinatubo eruption.

 

SDS seminar’s website: https://auburn.edu/cosam/datascienceseminar/index.htm

DMS Statistics and Data Science Seminar
Sep 24, 2025 01:00 PM
ZOOM


xiadong

Speaker: Xiaodong Li (University of California, Davis)

Title: Estimating SNR in High-Dimensional Linear Models: Robust REML and a Multivariate Method of Moments

 

Abstract: This talk presents two complementary approaches to estimating signal-to-noise ratios (and residual variances) in high-dimensional linear models, motivated by heritability analysis. First, I show that the REML estimator remains consistent and asymptotically normal under substantial model misspecification—fixed coefficients and heteroskedastic and possibly correlated errors. Second, I extend a method-of-moments framework to multivariate responses for both fixed- and random-effects models, deriving asymptotic distributions and heteroskedasticity-robust standard-error formulas. Simulations corroborate the theory and demonstrate strong finite-sample performance.

Host: Haoran Li

SDS seminar’s website: https://auburn.edu/cosam/datascienceseminar/index.htm

DMS Statistics and Data Science (SDS) Seminar
Sep 17, 2025 01:00 PM
352 Parker Hall


yin tang

Speaker: Yin Tang (University of Kentucky)

 Title: Belted and Ensembled Neural Network for Linear and Nonlinear Sufficient Dimension Reduction

 

Abstract: We introduce a unified, flexible, and easy-to-implement framework of sufficient dimension reduction that can accommodate both linear and nonlinear dimension reduction, and both the conditional distribution and the conditional mean as the targets of estimation. This unified framework is achieved by a specially structured neural network -- the Belted and Ensembled Neural Network (BENN) -- that consists of a narrow latent layer, which we call the belt, and a family of transformations of the response, which we call the ensemble. By strategically placing the belt at different layers of the neural network, we can achieve linear or nonlinear sufficient dimension reduction, and by choosing the appropriate transformation families, we can achieve dimension reduction for the conditional distribution or the conditional mean. Moreover, thanks to the advantage of the neural network, the method is very fast to compute, overcoming a computation bottleneck of the traditional sufficient dimension reduction estimators, which involves the inversion of a matrix of dimension either p or n. We develop the algorithm and convergence rate of our method, compare it with existing sufficient dimension reduction methods, and apply it to two data examples.

https://arxiv.org/abs/2412.08961

Host: Haotian Xu

SDS seminar’s website: https://auburn.edu/cosam/datascienceseminar/index.htm

DMS Statistics and Data Science Seminar
Apr 23, 2025 02:00 PM
354 Parker Hall


barrientos

Speaker: Dr. Andrés Felipe Barrientos (Assistant Professor, Department of Statistics, Florida State University)

Title: Bayesian nonparametric modeling of mixed-type bounded data

 

Abstract: We propose a Bayesian nonparametric model for mixed-type bounded data, where some variables are compositional and others are interval-bounded. Compositional variables are non-negative and sum to a given constant, such as the proportion of time an individual spends on different activities during the day or the fraction of different types of nutrients in a person's diet. Interval-bounded variables, on the other hand, are real numbers constrained by both a lower and an upper bound. Our approach relies on a novel class of random multivariate Bernstein polynomials, which induce a Dirichlet process mixture model of products of Dirichlet and beta densities. We study the theoretical properties of the model, including its topological support and posterior consistency. The model can be used for density and conditional density estimation, where both the response and predictors take values in the simplex space and/or hypercube. We illustrate the model's behavior through the analysis of simulated data and data from the 2005-2006 cycle of the U.S. National Health and Nutrition Examination Survey.

Joint work with Rufeng Liu, Claudia Wehrhahn, and Alejandro Jara.

DMS Statistics and Data Science Seminar
Apr 16, 2025 02:00 PM
354 Parker Hall


jiwonpark

Speaker: Dr. Jiwon Park (postdoctoral researcher, Department of Epidemiology, Johns Hopkins University)

 

Title: A Robust Pleiotropy Testing Method with Applications to Inflammatory Bowel Disease Subtypes with Sample Overlap

 

Abstract: Pleiotropy, where a genetic region influences multiple traits, is common in complex diseases and provides insight into shared biological mechanisms. However, identifying pleiotropic loci remains challenging, especially for correlated traits or case-control studies with overlapping samples. We present PLACO+, a statistical method for detecting pleiotropic associations using GWAS summary statistics from two traits. PLACO+ models a composite null hypothesis with an inflated variance structure, allowing for partial associations, and computes analytical p-values based on the distribution of the product of correlated Z-scores. Applied to genome-wide studies of inflammatory bowel disease (IBD) subtypes—Crohn’s disease and ulcerative colitis—PLACO+ identifies shared genetic loci missed by conventional approaches, particularly when effects are in opposite directions. These results demonstrate the utility of PLACO+ in uncovering novel pleiotropic signals in complex trait genetics.

More Events...