Tuesday, January 18, 2022
5:15 p.m. — room Jaures (29 rue d’Ulm, 75005 Paris)
Spin glasses are useless. Yet their study, triggered by pure intellectual interest, has created a formidable new branch of statistical physics distinguished this year by the Nobel prize attributed to Giorgio Parisi. Several decades after being deciphered, the spin glass mystery has found applications in many other fields, from protein folding to computational neurosciences, information theory, economics theory, signal processing or machine learning. This talk will explain how this “spin glass cornucopia” developed, and explain how some key concepts and methods have been used in other fields, taking specific examples in large-scale inference problems.