Daniel Tubbenhauer, "Mathematics, algebra and AI"Public Event
zur Anmeldung
Zusammenfassung:
The Ouroboros (the snake that eats itself) is a fitting image for this talk. I’ll explore how mathematics, especially linear and non-linear algebra, underpins the structure of neural networks in machine learning. At the same time, I’ll reflect (in a personal and informal way) on whether, how, and why (or why not) machine learning might be useful for advancing mathematics itself; particularly in algebra.
Daniel Tubbenhauer
Daniel is ARC Future Fellow at the University of Sydney, working in algebra and representation theory; especially categorical and diagrammatic approaches. Previously they held positions in Zürich, Bonn, and Aarhus. PhD from Göttingen in 2013 on categorification and its applications. They have co-organised various events in representation theory and categorification, and were involved in outreach through Spectra and WAMSOC. Daniel also runs a YouTube channel about mathematics, aiming to make abstract concepts more accessible and visual.
Stefan Hartmann