On July 5th, a group of our principal investigators, early career scientists, and members of the public gathered for another session on our SmallData Seminar Series. In this Bazaar edition, representatives from projects handling similar data structures presented some of their methodologies, showing insights into how they used some of their data analysis tools.
Jacques Hermes, a doctoral researcher of one of our PIs, Jens Timmer, began the session by talking about the challenges of Fisher information and the rise of profile likelihood, followed up by a research based guide on model reduction.
Our PI, Clemens Kreutz, presented a different approach to profile likelihood, showing how he used it on his research project to quantify prediction uncertainties.
Our doctoral researcher, Clemens Schächter, closed the introductory section by talking about model selection for disease trajectories and treatment effects in latent space by combining variational autoencoders with random effects models.
This was followed up by group sessions in which members interested in each of these forementioned topics had lively discussions on how such methodologies could be implemented in their own projects.
We are thrilled with the collaboration spirit shown and look forward to having you at our next Bazaar Seminar!
Institute of Medical Biometry and Statistics,
Faculty of Medicine and Medical Center –
University of Freiburg
