Qualitative Probabilistic Networks for Planning Under Uncertainty

TL;DR

This paper introduces a qualitative reasoning approach using probabilistic networks to support planning under uncertainty, providing valuable insights despite weaker conclusions than full probability models.

Contribution

It develops a qualitative algebra over Bayesian networks to infer action influences, aiding decision-making without requiring complete probability distributions.

Findings

Enables reasoning about action influence qualitatively

Helps eliminate inferior plans and identify tradeoffs

Provides explanations for probabilistic models

Abstract

Bayesian networks provide a probabilistic semantics for qualitative assertions about likelihood. A qualitative reasoner based on an algebra over these assertions can derive further conclusions about the influence of actions. While the conclusions are much weaker than those computed from complete probability distributions, they are still valuable for suggesting potential actions, eliminating obviously inferior plans, identifying important tradeoffs, and explaining probabilistic models.