A Theoretical Framework for Context-Sensitive Temporal Probability Model Construction with Application to Plan Projection

TL;DR

This paper introduces a context-sensitive temporal probability logic for Bayesian networks, enabling focused inference and efficient plan projection, with an application to cardiac treatment evaluation.

Contribution

It presents a novel logic framework and a Bayesian network construction algorithm that provides sound and complete answers for temporal probabilistic reasoning.

Findings

Algorithm successfully constructs Bayesian networks for complex queries.

Application demonstrates effectiveness in evaluating cardiac treatments.

Framework improves inference efficiency by focusing on relevant contexts.

Abstract

We define a context-sensitive temporal probability logic for representing classes of discrete-time temporal Bayesian networks. Context constraints allow inference to be focused on only the relevant portions of the probabilistic knowledge. We provide a declarative semantics for our language. We present a Bayesian network construction algorithm whose generated networks give sound and complete answers to queries. We use related concepts in logic programming to justify our approach. We have implemented a Bayesian network construction algorithm for a subset of the theory and demonstrate it's application to the problem of evaluating the effectiveness of treatments for acute cardiac conditions.