Cluster-Dags as Powerful Background Knowledge For Causal Discovery

TL;DR

This paper introduces Cluster-DAGs as a flexible prior knowledge framework to improve causal discovery in high-dimensional data, demonstrating superior performance of the modified algorithms over baselines.

Contribution

It proposes using Cluster-DAGs as a novel prior knowledge approach and develops two new algorithms, Cluster-PC and Cluster-FCI, for causal discovery.

Findings

Cluster-PC and Cluster-FCI outperform baseline methods on simulated data.

Cluster-DAGs provide greater flexibility than existing background knowledge methods.

Empirical results show improved accuracy in causal graph recovery.

Abstract

Finding cause-effect relationships is of key importance in science. Causal discovery aims to recover a graph from data that succinctly describes these cause-effect relationships. However, current methods face several challenges, especially when dealing with high-dimensional data and complex dependencies. Incorporating prior knowledge about the system can aid causal discovery. In this work, we leverage Cluster-DAGs as a prior knowledge framework to warm-start causal discovery. We show that Cluster-DAGs offer greater flexibility than existing approaches based on tiered background knowledge and introduce two modified constraint-based algorithms, Cluster-PC and Cluster-FCI, for causal discovery in the fully and partially observed setting, respectively. Empirical evaluation on simulated data demonstrates that Cluster-PC and Cluster-FCI outperform their respective baselines without prior knowledge.