On the Conditional Independence Implication Problem: A Lattice-Theoretic Approach

TL;DR

This paper introduces a lattice-theoretic framework for analyzing the conditional independence implication problem in discrete probability measures, providing a complete inference system and practical heuristics.

Contribution

It develops a novel lattice-based approach with a sound and complete inference system for various classes of CI statements, including saturated and stable cases.

Findings

A finite, sound, and complete inference system for CI statements.

Heuristics that approximate the implication criterion in polynomial time.

Experimental results comparing with existing inference algorithms.

Abstract

A lattice-theoretic framework is introduced that permits the study of the conditional independence (CI) implication problem relative to the class of discrete probability measures. Semi-lattices are associated with CI statements and a finite, sound and complete inference system relative to semi-lattice inclusions is presented. This system is shown to be (1) sound and complete for saturated CI statements, (2) complete for general CI statements, and (3) sound and complete for stable CI statements. These results yield a criterion that can be used to falsify instances of the implication problem and several heuristics are derived that approximate this "lattice-exclusion" criterion in polynomial time. Finally, we provide experimental results that relate our work to results obtained from other existing inference algorithms.