CAN7 seminar: Probabilistic modelling
Table of contents
Lecturer
PostDoc
NameDr. Sebastian Bitzer
Send encrypted mail via the SecureMail portal (for TUD external users only).
Chair of Cognitive Computational Neuroscience
Visiting address:
Falkenbrunnen, Raum 142 Chemnitzer Str. 46b
01187 Dresden
Schedule
Thursdays, 4. DS, FAL
Course Website
For more information, go to the course website
Description
Most information is uncertain, especially from the viewpoint of the brain which has to make sense of the environment based on a restricted set of sensory measurements. How do we cope with uncertainty, i.e., what is the best way to reason under uncertainty? Bayesian statistics provides a consistent formal framework to do this. Consequently, it has been suggested that the brain also implements some form of Bayesian inference. In the seminar you will learn the basics of Bayesian statistics and will see how it helps to identify problems with null hypothesis significance testing (why most published research findings are false). Furthermore, we will cover experimental evidence that people actually reason with uncertainty as predicted by Bayesian statistics. Finally, we will investigate when and how probabilistic models may help to understand activity in the brain. The seminar will mostly be based on presentations by students with intermittent explanations by myself. There is also the possibility of a small practical component.