Action Networks: A Framework for Reasoning about Actions and Change under Uncertainty

Moises Goldszmidt and Adnan Darwiche

This work proposes action networks as a semantically well founded framework for reasoning about actions under uncertainty. Action networks extend probabilistic causal networks to allow the representation of actions as directly controlling specific events in the domain (i.e., setting the value of nodes in the network) subject to preconditions. They also introduce models of time and persistence, and the capability of specifying uncertainty at different levels of abstraction. This paper describes both recent results and work in progress.

This page is copyrighted by AAAI. All rights reserved. Your use of this site constitutes acceptance of all of AAAI's terms and conditions and privacy policy.