USI - Email

Archive / INF Seminars / INF_2024_03_05_Joshua_Bon

USI - Email

	Bayesian score calibration for approximate models

Host: Prof. Antonietta Mira

Martedì

05.03

USI Campus Est, room D1.14, sector D // Online on Microsoft Teams

13:30 - 14:30


	Joshua Bon

Abstract:
Scientists continue to develop increasingly complex mechanistic models to reflect their knowledge more realistically. Statistical inference using these models can be challenging since the corresponding likelihood function is often intractable and model simulation may be computationally burdensome. Fortunately, in many of these situations, it is possible to adopt a surrogate model or approximate likelihood function. It may be convenient to conduct Bayesian inference directly with the surrogate, but this can result in bias and poor uncertainty quantification. In this paper we propose a new method for adjusting approximate posterior samples to reduce bias and produce more accurate uncertainty quantification. We do this by optimizing a transform of the approximate posterior that maximizes a scoring rule. Our approach requires only a (fixed) small number of complex model simulations and is numerically stable. We demonstrate good performance of the new method on several examples of increasing complexity. Joint work with: David J Warne, David J Nott, Christopher Drovandi.

Biography:
Dr Joshua Bon is a postdoctoral researcher working Université Paris Dauphine-PSL. He completed his PhD in 2022 in new methods for Bayesian statistical inference and computation with sequential Monte Carlo at the Queensland University of Technology (QUT) under supervision of Professor Christopher Drovandi (QUT) and Anthony Lee (University of Bristol). Since September 2023, Dr Bon has been working with Professor Christian Robert (Université Paris Dauphine-PSL) on principled Bayesian inference algorithms with privacy guarantees. Dr Bon’s core research is in computational statistics, focussing on developing and analysing Bayesian algorithms including sequential Monte Carlo and Markov chain Monte Carlo. In research led by Dr Bon, he specialises in the approximate computation of intractable models, developing new algorithms to understand real-world phenomena in situations where current methods are unsuitable or computationally infeasible. He also enjoys working with applied scientists to develop principled and robust data analysis for their data. This involves designing Bayesian models to incorporate domain and expert knowledge. He also develops open-source software for these projects. As a collaborator, he has provided his statistical expertise to colleagues in social science, public health, sports science, demography, and ecology.

Click here to join online