Séminaires : Combinatoire, Optimisation, et Interactions

The purpose of this seminar is to foster exchanges within the CO team of IMJ-PRG, and also with the surrounding scientific community. As such, its range of topics should be quite broad. We initially plan one or two sessions per month.

Ce séminaire a pour but de développer les échanges au sein de l'équipe Combinatoire et Optimisation, et avec la communauté scientifique environnante. Ses thèmes seront donc larges. Nous prévoyons un rythme initial de une à deux séances par mois.

We study the implications of the « no-regret property » for  procedures in continuous and discrete time used in  on-line learning, game theory dynamics and convex optimization. 
We will recall the origin of the problem  and describe several ramifications (calibration, variational inequalities).