Future events
Mostly Monte Carlo
24/04/24
MCMC when you do not want to evaluate the target distribution
Guanyang Wang Rutgers University
In sampling tasks, it is common for target distributions to be known up to a normalizing constant. However, in many situations, evaluating even the unnormalized distribution can be costly or infeasible. This issue arises in scenarios such as sampling from the Bayesian posterior for large datasets and the ‘doubly intractable’ distributions. We provide a way to unify various MCMC algorithms, including several minibatch MCMC algorithms and the exchange algorithm. This framework not only simplifies the theoretical analysis of existing algorithms but also creates new algorithms. Similar frameworks exist in the literature, but they concentrate on different objectives
07.07.2024 ISBA-Cini : Artificial intelligence, data sharing and regulation
Ocean co-organizes with ISBA the conference « Artificial Intelligence, Data Sharing, and Regulation » at Venice. The focus will be the most recent advances and the future perspectives of research in artificial intelligence and statistics, putting particular emphasis on ethical and regulatory issues. Christian Robert, Stéphanie Allassonnière and Michael Jordan will speak on Statistics and Machine Learning as well as their integration in the European Market.
07.2024 First Ocean summer retreat at the Ca’ Foscari University of Venice
Ocean will have its first Summer retreat from the 7th to the 12th of July, 2024 at the Ca’ Foscari University of Venice! This event will bring together Ocean researchers and students to make waves in the world of science and innovation! We are planning talks, working sessions to fuel collaboration, and short courses exploring the intersection of machine learning and Economics.
Past events
03.2024 Seminar at the Houches
In March 2024, Ocean held its first internal seminar in the Houches, France. This meeting saw some Ocean researchers and invited experts (Arnoldo Frigessi, Steve MacEachern, Sylvia Richardson, Judith Rousseau) gather with short presentations and joint work sessions on open problems and potential collaborations.
The scientific program was focused on two core research axes of Ocean: differential privacy in collaborative models and Bayesian inference and sampling.
03.2024 Launch of the Ocean seminar
The first session of the Ocean seminar was held on the 1st of March! This seminar aims at gathering all Ocean researchers and students once a month to foster collaboration and discuss advances about ongoing research. Antoine Scheid inaugurate dthis seminar by presenting his work about incentives in a principal-agent bandit framework.
02.2024 Warwick wired to Paris
Stanislas Duché and Christian Robert spent a week at Warwick University to collaborate with Gareth Roberts‘ team about privacy-related projects. This exchange was ponctuated by very interesting discussions with Gareth Roberts,, Ritabrata Dutto, Massimiliano Tamborini, and their students, and a presentation of Stanislas to Warwick’s PhD reading group.
01.2024 From Warwick to All-About-That-Bayes
On the 23th of January 2024 at PariSanté Campus, Ritabrata Dutta (Warwick University) introduced « Bayesian Model Averaging with exact inference of likelihood- free Scoring Rule Posteriors » at the All-About-That-Bayes seminar. This seminar, held at OCEAN’s PariSanté campus offices, brought together teams from Warwick and Dauphine as well as Paris statisticians.
12.2023 Ocean was at the CIRM!
Last week, the CIRM in Marseille hosted the seminar ‘From Matching to Markets: a tale of Mathematics‘, with the aim of gathering researchers from the fields of mathematics, computer science, and economics to study of matching problems and the design of their markets.
Ocean was well represented, since Vianney Perchet was a member of the Organizing Committee and Michael I. Jordan part of the Scientific Committee. Michael gave a talk about Statistical Contract Theory while Etienne Boursier took the floor to present an overview on Multiplayer Bandits. We learnt a lot there and had fun, thank you to everyone for this event!
OCEAN 