Summary: Appears in Proceedings of 18th Int'l Joint Conference on Artificial Intelligence (IJCAI '03).
Logical Filtering
Eyal Amir and Stuart Russell
Computer Science Division, University of California at Berkeley
Berkeley, CA 94720­1776, USA
{eyal,russell}@cs.berkeley.edu
Abstract
Filtering denotes any method whereby an agent up­
dates its belief state---its knowledge of the state of
the world---from a sequence of actions and obser­
vations. In logical filtering, the belief state is a log­
ical formula describing possible world states and
the agent has a (possibly nondeterministic) logi­
cal model of its environment and sensors. This
paper presents efficient logical filtering algorithms
that maintain a compact belief state representa­
tion indefinitely, for a broad range of environment
classes including nondeterministic, partially ob­
servable STRIPS environments and environments
in which actions permute the state space. Efficient

Source: Amir, Eyal - Department of Computer Science, University of Illinois at Urbana-Champaign

Collections: Computer Technologies and Information Sciences