Vinayak Rao

I am an Associate Professor at the Department of Statistics at Purdue University, with a courtesy appointment in Computer Science. Before coming to Purdue, I was a postdoc at Duke University working with David Dunson. I obtained my PhD from the Gatsby Unit at the University College London, working with Yee Whye Teh.

Broadly, I am interested in how ideas and tools from statistics, computation and stochastic processes can help develop statistical models that are flexible and robust, and algorithms that are efficient. Particular areas of interest include Bayesian nonparametrics, generative models, continuous-time models, Markov chain Monte Carlo and variational inference.

Select recent preprints/publications:

• Pei, X., Kim, M. and Rao, V.A. (2026)
  Exact Gibbs sampling for stochastic differential equations with gradient drift and constant diffusion. [arxiv:2602.00512]
  
  
• Banerjee, I., Honnappa, H., and Rao, V.A. (2025)
  Offline Estimation of Controlled Markov Chains: Minimaxity and Sample Complexity. Operations Research [abs]
  
  
• Beraha, M., Favaro, S. and Rao, V.A. (2024)
  MCMC for Bayesian nonparametric mixture modeling under differential privacy. J. Comput. Graph. Stat. [pdf]
  
  
• Awan, J., and Rao, V.A. (2023)
  Privacy-aware rejection sampling. J. of Mach. Learn. Res. [abs]
  
  
• Ju, N., Awan, J., Gu R. and Rao, V.A. (2022)
  Data augmentation MCMC for bayesian inference from privatized data. NeurIPS [abs]