A Probabilistic Network of Predicates

TL;DR

This paper introduces a probabilistic network model where nodes are unary predicates, allowing cycles and more flexible domain knowledge representation compared to traditional Bayesian networks.

Contribution

It proposes a novel probabilistic network framework with predicate nodes and cycles, overcoming limitations of propositional and acyclic Bayesian networks.

Findings

Supports cyclic causal reasoning

Handles recursive plans

Enables static domain knowledge representation

Abstract

Bayesian networks are directed acyclic graphs representing independence relationships among a set of random variables. A random variable can be regarded as a set of exhaustive and mutually exclusive propositions. We argue that there are several drawbacks resulting from the propositional nature and acyclic structure of Bayesian networks. To remedy these shortcomings, we propose a probabilistic network where nodes represent unary predicates and which may contain directed cycles. The proposed representation allows us to represent domain knowledge in a single static network even though we cannot determine the instantiations of the predicates before hand. The ability to deal with cycles also enables us to handle cyclic causal tendencies and to recognize recursive plans.