STAT@UVM Colloquium Series: Nianqiao (Phyllis) Ju

About this event

Join the STAT@UVM colloquium series for talks connecting statistics, data science, and applications across disciplines.

Bayesian Inference from Privatized Data via Data Augmentation

with Nianqiao (Phyllis) Ju

Dartmouth College

ABSTRACT: Analyzing data under privacy constraints is like dancing with shackles, as privacy protection requires modified analysis procedures and inevitably reduces data utility. Performing statistical inference on privatized data is even more challenging, as it requires deriving reliable insights from lower-quality data. Drawing on three recent studies, this talk presents a data augmentation Markov Chain Monte Carlo framework for Bayesian inference from privatized data. The core idea is to treat the unobserved confidential data as latent variables. We discuss an independent Metropolis-within-Gibbs sampler and a recent algorithmic improvement. We present spectral gap bounds of reversible hybrid Gibbs samplers, validating the convergence of our data augmentation algorithms.

BIO: Nianqiao (Phyllis) Ju is an Assistant Professor of Mathematics at Dartmouth College. Her research spans Bayesian statistics, Monte Carlo methods, differential privacy, and applied statistics, with applications in privacy-aware data analysis and infectious disease modeling. Before joining Dartmouth, she was an Assistant Professor of Statistics at Purdue University. She holds a Ph.D. in Statistics from Harvard University (2021), where she was advised by Pierre Jacob, and a B.A. in Mathematics and Physics from Wellesley College (2016). She received a 2025 Blackwell-Rosenbluth Award from the International Society for Bayesian Analysis, which recognizes outstanding junior Bayesian researchers.