Lecture to discuss inference methodologies in infinite-dimensional statistical models
Statistical models are becoming increasingly complex every day. In particular, it has become common to consider infinite-dimensional models, in which (augmented) sample space and parameter space have infinite dimensions. To debate this subject, the FGV´s School of Applied Mathematics (FGV EMAp) will hold a seminar titled “The Magic of Retrospective Sampling: tractable exact solutions for intractable infinite-dimensional problems” on August 22. The event will take place in Auditorium 537 in FGV’s main building, at Praia de Botafogo 190, Botafogo, Rio de Janeiro, starting at 4 pm.
To talk about this topic, the guest will be Flávio Bambirra Gonçalves, a professor at Minas Gerais Federal University, who has a PhD in statistics from the University of Warwick in the United Kingdom. He will discuss the serious intractability problems that these models inevitably lead to in a context of inference that until recently required the use of finite-dimensional approximations, which are a considerable source of error and often difficult to quantify and control.
The lecture will present the idea of retrospective sampling and discuss some general methodologies for making exact inferences in infinite-dimensional models. By way of illustration, the guest will use examples related to Cox processes and SDE-driven models.
For more information and to sign up for the talk, visit this website.