Speaker: Gerhard Lakemeyer
Department of Computer Science, Aachen University of Technology
Title: Progressing a knowledge base under uncertainty
In planning and action languages the notion of progressing an initial knowledge base when actions are performed plays an important role. While progression is difficult in general, a number of restrictions, which are both computationally well behaved and of practical relevance, have been identified over the past years. However, very little is known yet about progression when actions have uncertain effects. In this talk I will demonstrate how such noisy actions can be dealt with in the framework of the situation calculus and knowledge bases with probabilistic beliefs.