Compositional Structures in Probabilistic Modeling (CSPM)

Scope

The scope of the seminar encompasses compositional structures such as, but not limited to, graphs, hypergraphs, posets, and their representations (functors, sheaves) but also operads and their use for and interaction with probabilistic modeling, for example, in machine learning, deep learning, reinforcement learning, multi-agent systems, games, etc. The aim of the seminar is to stay updated on novel findings around modeling, theoretical developments, and algorithms for compositional structures in probabilistic modeling.

To join the mailing list or the group of the seminar please send a mail to compositional-structures-in-probabilistic-modeling@googlegroups.com

We will try our best to record the session, and the videos will be posted on our YouTube channel: Visit Our YouTube Channel

If you are interested in speaking at the seminar, please send me a message to gregoireserper 'at' gmail 'dot' com

Sessions

  1. Date: April 5, 2024
    Time: 4:00 PM (Paris time)
    Topic: Sheaves in data science
    Speaker: Gregoire Sergeant-Perthuis (Sorbonne Universite)
    Where: salle 15-16-411, IMJ-PRG, Sorbonne Université, 4 Place Jussieu
    Video of the session

  2. Date: 29 May, 2024
    Time: 6:00 PM (Paris time)
    Topic: Active Inference in String Diagrams
    Speakers: Sean Tull (Cambridge Quantum Computing and Topos Institute)
    Where: Room 1516-4-11, IMJ-PRG, Sorbonne Université, 4 Place Jussieu

Organizers

  • Gregoire Sergeant-Perthuis (Sorbonne Universite)