Dr Shenggang Hu PhD

• PhD in Statistics, University of Essex, 2023
• MSc in Robotics and Computation, University College London, 2019
• MMath in Mathematics, University of Oxford, 2018
• BA in Mathematics, University of Oxford, 2017

Shenggang Hu completed a BA and MMath in Mathematics at the University of Oxford, followed by an MSc in Robotics and Computation at University College London. He completed his PhD in Statistics at the University of Essex under supervision of Professor Hongsheng Dai in 2023. His doctoral research developed Monte Carlo methods for exact sampling, density fusion and statistical disaggregation under constraints.

During his PhD, Shenggang taught in the Department of Mathematical Sciences at Essex, first as a Graduate Teaching Assistant and then as a Fixed Term Lecturer. He designed and delivered lectures for “Statistical Methods” and classes for other modules.

From 2023 to 2026 he was a Postdoctoral Researcher in the Department of Statistics at the University of Warwick. His postdoctoral research focused on differential privacy and Monte Carlo methods. He joined the School of Mathematics at University of Birmingham in 2026 as an Assistant Professor.

• Foundation and Calculus
• Programming in Python

Shenggang's research is at the interface of Bayesian statistics, computational statistics and privacy-preserving inference. He develops Monte Carlo methodology with theoretical guarantees, including exact and constrained sampling, Bernoulli factories, and MCMC methods.

His recent work studies differential privacy for Bayesian posterior sampling, particularly under contaminated and robust statistical models. This research examines how model robustness can control sensitivity to individual observations and how privacy, statistical efficiency and computational performance interact as sample size grows.

His doctoral research focused on rejection sampling, density fusion and constrained disaggregation. His broader interests include MCMC scaling limits, robust Bayesian inference, online Bayesian learning and the responsible use of statistical methods with sensitive data.