Dr Nicola Branchini
Hello! I am a Research Fellow in the Department of Statistics at the University of Warwick, working with Gareth Roberts under the ProbAI Hub.
Please see also my personal webpage. Link opens in a new window
My real interests are broad, spanning computational statistics and statistical/probabilistic machine learning, with a focus on methodology. For my PhD, I have been focusing on developing methodology in Monte Carlo.
A non-comprehensive sample of some specific interests:
- Importance sampling–
I'm interested in all kinds of aspects of importance sampling, such as dimension scaling, good choices of proposal densities, combinations with MCMC, robustness to inexact function evaluations, and practical diagnostics
- Bayesian cross-validation– The Bayesian counterpart to classical cross-validation presents an interesting setup for IS and computational methods, and I'm interested in extensions and theory around Bayesian CV.
- Heavy tails– Approximating distributions with heavy tails is challenging and still a very active research area. I'm also interested in using concepts from heavy tails to understand the properties of MC estimators.
- …and adjacent ideas– I like to keep learning new concepts from neighbouring fields. For instance, at the moment I am curious about how we can use concepts from Monte Carlo to understand properties of generative (agentic?) models and make them more reliable.
Please do not hesitate to contact me to talk about research.