Western Michigan University Department of Statistics

03/25/2026

Today is Giving Day at WMU! You can donate to the Department of Statistics at the link below. Donations can be made through tomorrow, Thursday, March 26th at 8:00 am.

Join me and make a gift to Western Michigan on Giving Day!

03/16/2026

March 25th is Giving Day at WMU! You can donate to the Statistics Department at the link below.

Join me and make a gift to Western Michigan on Giving Day!

03/05/2026

Tomorrow, Friday, March 6th is Spirit Day and next week, March 9th through 13th, is spring break. Enjoy your time off and be safe!

02/26/2026

Colloquium Presentation by Ruyu Zhou on Friday, February 27th 11:00 am - 12:00 pm in the Alavi Commons Rooms (6625 Everett Tower).

Title: Differential Privacy and Statistics: Privatized Inference and the Inherent Privacy of Sampling

Abstract: Privacy-preserving data analysis has become a central challenge in modern statistics, with Differential Privacy (DP) emerging as the gold standard for protecting individual-level information. In this talk, I will present two projects at the intersection of DP and statistics. First, focusing on privatized inference, I will introduce a general framework for privacy-preserving statistical inference that constructs privatized interval estimators via consistent, privatized posterior quantile estimation. I theoretically establish mean-squared error consistency for the proposed estimators and demonstrate improved privacy-utility tradeoffs through extensive empirical experiments. Second, focusing on the inherent privacy guarantees provided by posterior sampling, I develop a unified Rényi divergence framework to quantify the DP guarantees achieved "for free" when releasing a single posterior sample. These theoretical results substantially tighten existing conservative bounds and broaden the class of Bayesian models for which inherent DP guarantees can be rigorously characterized.

02/24/2026

Colloquium Presentation by Dr. Xiaomeng Ju on Wednesday, February 25th 11:00 am - 12:00 pm in the Alavi Commons Rooms (6625 Everett Tower)

Title: Bayesian Modeling for Functional and Matrix Data with Applications to Neuroimaging Analysis

Abstract: Neuroimaging data present fundamental statistical challenges: they are high-dimensional and exhibit complex structures. In this talk, I will present Bayesian methods developed for functional data and matrix-valued data motivated by neuroimaging applications, emphasizing interpretability, scalability, and uncertainty quantification.
I will first introduce Bayesian methods developed for two types of functional data derived from evoked electroencephalogram (EEG) signals in multi-condition settings, including (1) time-frequency representations and (2) dynamic functional connectivity. The proposed models jointly account for the data’s multilevel structure, functional nature, and subject-level covariates by incorporating covariate-dependent fixed effects and multilevel random effects. These methods are evaluated through simulations and applied to EEG data examining the effects of alcoholism on cognitive processing in response to visual stimuli. I will then introduce Bayesian methods developed for matrix-valued data, with applications to predicting scalar outcomes using resting-state functional magnetic resonance imaging (fMRI) functional connectivity. Lastly, I will conclude by discussing future directions for developing structure-aware statistical models for neuroimaging data.

02/23/2026

Colloquium Presentation by Shanshan Wang on Monday, February 23rd 11:00 am - 12:00 pm in the Alavi Commons Rooms (6625 Everett Tower)

Title: Dimension Reduction for Conditional Quantiles of Functional Data with Applications

Abstract: Functional data analysis holds transformative potential across many fields but has largely focused on mean regression, with relatively limited attention to quantile regression. Moreover, the infinite-dimensional nature of functional predictors necessitates effective dimension reduction techniques. In this talk, I present a unified framework for sufficient dimension reduction in conditional quantile regression with functional data. The framework includes both linear and nonlinear dimension reduction methods, as well as extensions that accommodate quantitative and categorical predictors simultaneously. I establish theoretical guarantees for the proposed approaches and demonstrate their practical performance through simulations and applications to health-related data, including fMRI studies and Parkinson’s disease.

02/14/2026

Colloquium Presentation by Dr. Qi Wang on Monday, February 16th 11:00 am - 12:00 pm in the Alavi Commons Rooms (6625 Everett Tower)

Title:
Modern Statistical Approaches to Modeling Spatial and Spatio-Temporal Processes

Abstract:
Deep learning methods have become increasingly popular for modeling complex, high-dimensional data. This work presents a set of projects that bridge traditional statistical modeling and modern deep learning for spatial and spatio-temporal data. The proposed approaches combine Bayesian hierarchical modeling with state-space and neural network representations to capture complex and non-linear dependence structures across space and time. The projects include pure Bayesian spatio-temporal Markov models for global snow cover dynamics, hierarchical echo state network models for dependent count time series, and spatial deep convolutional neural networks for prediction for spatial dataset at unobserved locations. Across multiple real and simulated datasets, these models demonstrate strong predictive performance and well-calibrated uncertainty.

11/30/2025

11/25/2025

Thanksgiving break starts at noon tomorrow, Wednesday, November 26th. Classes will resume as normal on Monday, December 1st.