Decision Theoretic Approaches for Focussed Bayesian Fusion

Conference paper

Links:
Authors:

Jennifer Sander
Jürgen Beyerer

Source:

Hans-Ulrich Heiß, Peter Pepper, Holger Schlinghoff, Jörg Schneider (eds.), INFORMATIK 2011 - Informatik schafft Communities, Gesellschaft für Informatik, 2011.

Pages:

478

Conference:

INFORMATIK 2011 - the 41th Annual Conference of the Gesellschaft für Informatik e.V. (GI), 6th Workshop Sensor Data Fusion: Trends, Solutions, Applications (SDF), Berlin, October 6 - 7, 2011

Focussed Bayesian fusion is a local Bayesian fusion technique by that high costs caused by Bayesian fusion can get circumvented. This publication addresses globally optimal decision making on the basis of a focussed Bayesian model. Therefore, common decision criteria under linear partial information and in particular principles of lazy decision making are applied. We also present an interval scheme for global posterior probabilities whose informativeness is notably high.