Associate Professor · Statistics · Virginia Tech
Jyotishka Datta
Bayesian methodology and theory for structured high-dimensional data — shrinkage, sparsity, and their applications.
About
I am an Associate Professor of Statistics at Virginia Tech (previously Assistant Professor here, and at the University of Arkansas, Fayetteville, 2016–2020). My research focuses on Bayesian methodology and theory for structured high-dimensional data: shrinkage estimation, sparse signal recovery, graphical modelling, and nonparametric Bayes.
Application domains include astronomy, cancer genomics, neuroscience, ecology, and crime forecasting.
I received the NSF CAREER Award in 2025 (DMS-2443282) and the Dayanand Naik Award from the Virginia Chapter of the ASA in 2023.
I don't maintain this page as frequently as I should. Google Scholar is the best source for my papers; for everything else, email me.
Professional preparation
-
Postdoctoral Fellow
Duke University (Statistics) & SAMSI · 2014–2016
Mentors: David B. Dunson & Sandeep S. Dave -
Ph.D. in Statistics
Purdue University · 2009–2014
Advisors: Jayanta K. Ghosh & Michael Yu Zhu - B.Stat & M.Stat Indian Statistical Institute · 2003–2008
Research & Publications
For a complete and up-to-date list see Google Scholar or my CV. Papers co-authored with Prof. Nick Polson carry an alphabetical author list.
Under review
- Polynomial Log-Marginals and Tweedie's Formula: When Is Bayes Possible? arXiv:2509.05823
- Prediction-Powered Inference with Inverse Probability Weighting arXiv:2508.10149
- Horseshoe-type Priors for Independent Component Estimation arXiv:2406.17058
- The Curious Problem of the Normal Inverse Mean: Robustness and Shrinkage arXiv:2410.20641
- Inverse Probability Weighting: from Survey Sampling to Evidence Estimation arXiv:2204.14121 · minor revision, NEJSDS
- On Posterior Consistency of Bayesian Changepoint Models (with N. Guha) arXiv:2102.12938
Selected publications
- Asymptotic Properties of Bayes Risk for the Horseshoe Prior Bayesian Analysis, 2013
- Bayesian Inference on Quasi-Sparse Count Data Biometrika, 2016
- Default Bayesian Analysis with Global-Local Shrinkage Priors Biometrika, 2016
- Lasso Meets Horseshoe Statistical Science, 2019
- Prediction Risk for Global-Local Shrinkage Regression JMLR, 2019
- Group Inverse-Gamma Gamma Shrinkage for Sparse Linear Models Bayesian Analysis, 2024
- Evidence Estimation in Gaussian Graphical Models JMLR, 2024
- Nonparametric Bayes Multiresolution Testing for Massive-Dimensional Rare Events Journal of Nonparametric Statistics, 2024
Awards & Funding
- 2025NSF CAREER Award — "Heavy-Tailed Priors for Robust Bayesian Inference in Ecology, Machine Learning, and Astronomy." DMS-2443282 · $450,000 (2025–2030). nsf.gov ↗
- 2023Dayanand Naik Award — Virginia Chapter of the American Statistical Association, for outstanding research and service to Virginia.
- 2020NSF-DMS 2015460 (Co-PI with N. Guha) — "New Directions in Bayesian Change-Point Analysis." $139,984 (2020–2023).
- 2018Robert and Sandra Connor Endowed Faculty Fellowship, University of Arkansas.
- 2013Honorable Mention, Best Theoretical Poster — O'Bayes 2013, Durham, NC.
- 2013William J. Studden Publication Award — Department of Statistics, Purdue University.
Teaching
- STAT 3504 Nonparametric Statistics Fall 2025
- CMDA 2006 Integrated Quantitative Sciences II Spring 2026
- STAT 4504/5504G Applied Multivariate Statistics
- CMDA 4654 Intermediate Data Analytics & Machine Learning
- STAT 5525 Data Analytics
- CMDA 2014 Data Matters
Contact
Address
Department of StatisticsVirginia Tech, Blacksburg, VA