New Rules for Causal Identification with Background Knowledge

TL;DR

This paper introduces two new rules for incorporating background knowledge into causal identification, improving efficiency and applicability in the presence of latent variables and observational data.

Contribution

The paper proposes novel rules for causal identification with background knowledge, reducing computational complexity and extending applicability to typical causality tasks.

Findings

Rules are applicable in determining possible causal effects.

Method circumvents exponential enumeration of block sets.

Enhances existing causal inference techniques.

Abstract

Identifying causal relations is crucial for a variety of downstream tasks. In additional to observational data, background knowledge (BK), which could be attained from human expertise or experiments, is usually introduced for uncovering causal relations. This raises an open problem that in the presence of latent variables, what causal relations are identifiable from observational data and BK. In this paper, we propose two novel rules for incorporating BK, which offer a new perspective to the open problem. In addition, we show that these rules are applicable in some typical causality tasks, such as determining the set of possible causal effects with observational data. Our rule-based approach enhances the state-of-the-art method by circumventing a process of enumerating block sets that would otherwise take exponential complexity.