The lunch starts at 12:00, and the talk will start around 12:20.
NB: Wednesday BigInsight Lunches are open to staff and students from any of the BigInsight partners, including UiO, but not to others.
Speakers: Helge Langseth, Professor in Computer Science, NTNU.
Location: 8. floor Niels Henrik Abel’s building.
Title: Deep learning or probabilities — or both?
Abstract: Motivated by a recent discussion we’ve had here (NTNU, red. anm.) I was thinking I’d like to talk about something in the intersection between deep learning and probabilistic models. Basically, I have come into this “old man-mode”, where I have started to insist that everything was better when **I** was young, and I have had these long rants about why we need probabilistic models, why deep learning cannot (on its own) be the answer, and how deep Bayesian models may strike a good balance between probabilistic soundness and modelling flexibility. I think such models are very relevant for us at our lab at least, as they may be one (if not *the*) solution both for interpretability (XAI-related argument), robustness (statistical/computational learning theory related), and for the actual decision scenarios that are given by use-cases we have been working with up here (uncertainty aware/introspective related).