Virginia Tech® home

Colloquium

  • Colloquium
  • STAT 5924, CRN 20257
  • Fridays
  • 2:30 pm to 3:30 pm
  • 300 Seitz

Colloquium Schedule Spring 2026

Shounak Chattopadhyay

An assistant Professor in the Department of Statistics, University of Virginia. Prior to joining UVA, I was a Postdoctoral Scholar at the University of California, Los Angeles, working with Dr. Marc A. Suchard. I completed my Ph. D. at the Department of Statistical Science, Duke University, under the supervision of Dr. David B. Dunson. Before Duke, I completed my Bachelors and Masters degrees in Statistics from the Indian Statistical Institute, Kolkata.

https://shounakch.github.io/

Title: Blessing of dimension in Bayesian inference on covariance matrices

Abstract: Bayesian factor analysis is routinely used for dimensionality reduction in modeling of high-dimensional covariance matrices. Factor analytic decompositions express the covariance as a sum of a low rank and diagonal matrix. In practice, Gibbs sampling algorithms are typically used for posterior computation, alternating between updating the latent factors, loadings, and residual variances. In this article, we exploit a blessing of dimensionality to develop a provably accurate posterior approximation for the covariance matrix that bypasses the need for Gibbs or other variants of Markov chain Monte Carlo sampling. Our proposed Factor Analysis with BLEssing of dimensionality (FABLE) approach relies on a first-stage singular value decomposition (SVD) to estimate the latent factors, and then defines a jointly conjugate prior for the loadings and residual variances. The accuracy of the resulting posterior approximation for the covariance improves with increasing samples as well as increasing dimensionality. We show that FABLE has excellent performance in high-dimensional covariance matrix estimation, including producing well-calibrated credible intervals, both theoretically and through simulation experiments. We also demonstrate the strength of our approach in terms of accurate inference and computational efficiency by applying it to a gene expression dataset.

Anne van Delft

About: Anne van Delft is a Tenure Track Assistant Professor in the Department of Statistics at Columbia University. She obtained a PhD at Maastricht University, the Netherlands, in December 2016. Before joining the Department of Statistics at Columbia University, she held a postdoctoral position in mathematical statistics at the Ruhr University in Bochum, Germany. 

Research interests: Anne van Delft's primary research interests lie in the area of stochastic processes that take values in function spaces, and in particular in the development of theory and methodology for function-valued time series with time-dependent characteristics. This is a line of research which she started to develop during her PhD, and is concerned with the analysis of sequential collections of data points that themselves come in the form of complex mathematical structures, such as curves, surfaces or manifolds. Inference techniques to analyze such data not only require a mathematically rigorous and quantitative study of their `shape' and dependence structure, but moreover must translate into computationally efficient methods. Examples can be found in (neuro-)imaging, climatology, genomics, and econometrics. She is especially interested in the development of appropriate statistical theory to further advance inference methods in these essential applications, which are characterized by dependence over time and space, and of which the dependence structure is of an evolutionary nature. 

https://sites.google.com/view/anne-van-delft/home/bio

Download Schedules from Colloquia Gone by

Contact Information

Department of Statistics (MC0439)
Hutcheson Hall, RM 406-A, Virginia Tech
250 Drillfield Drive
Blacksburg, VA 24061

Phone: 540-231-5657

Department Head:
Robert B. Gramacy