28/04/2026
La scorsa settimana abbiamo avuto il piacere di ospitare il Prof. Stephen Walker, Professore Onorario presso l’University of Kent (UK), che ha tenuto un corso rivolto a dottorandi e dottorande e un seminario aperto a tutta la comunità accademica.
Il suo intervento ha approfondito il tema dell’incertezza in ambito Bayesiano, offrendo una prospettiva sulle possibili strategie di quantificazione.
Ringraziamo sentitamente il Prof. Walker per la disponibilità e per il contributo offerto alla nostra comunità.
Di seguito l’abstract del seminario:
“The talk will discuss the source of Bayesian uncertainty. Statisticians typically observe a part of the data of what would need to be observed for no uncertainty to exist. The missing part of the data represents the source of uncertainty. In order to quantify this uncertainty a model for the missing data conditional on the observed data is required to be constructed. Classical Bayesian approaches do this through a prior and posterior. Alternative strategies based on one step ahead predictive densities or density estimators are also possible. A key tool for the construction is a martingale.”
______________________
Last week we had the pleasure of hosting Prof. Stephen Walker, Honorary Professor at the University of Kent (UK), who delivered a course for PhD students as well as an open seminar for the wider academic community.
His talk focused on the nature of Bayesian uncertainty, providing a clear and insightful perspective on different approaches to its quantification.
We are sincerely grateful to Prof. Walker for his time and for the valuable contribution to our community.
