Identification in Tree-shaped Linear Structural Causal Models

TL;DR

This paper introduces a novel method for identifying causal parameters in tree-shaped linear structural causal models by leveraging missing cycles of bidirected edges, offering an efficient alternative to Gr"obner basis approaches.

Contribution

The paper presents a new identification technique using missing cycles in tree-shaped models, reducing computational complexity compared to existing algebraic methods.

Findings

Identifies causal parameters using missing cycles of bidirected edges.

Provides explicit solutions to quadratic equations for causal parameters.

Develops an efficient algorithm for model identification.

Abstract

Linear structural equation models represent direct causal effects as directed edges and confounding factors as bidirected edges. An open problem is to identify the causal parameters from correlations between the nodes. We investigate models, whose directed component forms a tree, and show that there, besides classical instrumental variables, missing cycles of bidirected edges can be used to identify the model. They can yield systems of quadratic equations that we explicitly solve to obtain one or two solutions for the causal parameters of adjacent directed edges. We show how multiple missing cycles can be combined to obtain a unique solution. This results in an algorithm that can identify instances that previously required approaches based on Gr\"obner bases, which have doubly-exponential time complexity in the number of structural parameters.