Workshop organized by Dr. Andreas Döpp (CAS Young Center/LMU), Tom Sterkenburg, Ph.D. (CAS Young Center/LMU).
From the initial design of experiments to the interactive optimization of systems, Bayesian action policies have demonstrated remarkable efficiency. Yet as we scale these systems up, various tensions arise. Principled solutions often become intractable, and the temptation of ad hoc heuristics grows. Starting each problem from scratch is no longer affordable, so actors need to accumulate experience across tasks through ever more complex, learned priors. At the same time, wider deployment entails new and uncertain settings in which models and objective functions may no longer be adequate.
Participants include the members of the CAS Research Focus “Bayesian Reasoning: A Grammar of Science?”.
Anmelde-Info
Für die Teilnahme ist eine Anmeldung unter info@cas.lmu.de erforderlich.