Structural intervention distance for evaluating causal graphs

DOI10.1162/NECO_A_00708zbMATH Open1414.05094arXiv1306.1043OpenAlexW2116993992WikidataQ50974236 ScholiaQ50974236MaRDI QIDQ5380224FDOQ5380224

Authors: Jonas Peters, Peter Bühlmann 

Publication date: 4 June 2019

Published in: Neural Computation (Search for Journal in Brave)

Abstract: Causal inference relies on the structure of a graph, often a directed acyclic graph (DAG). Different graphs may result in different causal inference statements and different intervention distributions. To quantify such differences, we propose a (pre-) distance between DAGs, the structural intervention distance (SID). The SID is based on a graphical criterion only and quantifies the closeness between two DAGs in terms of their corresponding causal inference statements. It is therefore well-suited for evaluating graphs that are used for computing interventions. Instead of DAGs it is also possible to compare CPDAGs, completed partially directed acyclic graphs that represent Markov equivalence classes. Since it differs significantly from the popular Structural Hamming Distance (SHD), the SID constitutes a valuable additional measure. We discuss properties of this distance and provide an efficient implementation with software code available on the first author's homepage (an R package is under construction).

Full work available at URL: https://arxiv.org/abs/1306.1043

Recommendations

• Characterization and greedy learning of interventional Markov equivalence classes of directed acyclic graphs
• Causal inference from graphical models
• Estimating high-dimensional intervention effects from observational data
• Causal reasoning with ancestral graphs
• Quantifying causal influences

zbMATH Keywords

directed acyclic graphCausal inference

Mathematics Subject Classification ID

Directed graphs (digraphs), tournaments (05C20) Graph theory (including graph drawing) in computer science (68R10) Planar graphs; geometric and topological aspects of graph theory (05C10) Distance in graphs (05C12)

Cites Work

• Causation, prediction, and search
• A characterization of Markov equivalence classes for acyclic digraphs
• Title not available (Why is that?)
• The max-min hill-climbing Bayesian network structure learning algorithm
• Causality. Models, reasoning, and inference
• Probabilistic graphical models.
• Ancestral graph Markov models.
• Estimating the effect of joint interventions from observational data in sparse high-dimensional settings
• Title not available (Why is that?)
• Learning high-dimensional directed acyclic graphs with latent and selection variables
• Identifiability of Gaussian structural equation models with equal error variances
• Matrix multiplication via arithmetic progressions
• 10.1162/153244303321897717
• CAM: causal additive models, high-dimensional order search and penalized regression

Cited In (7)

• Causal discovery in heavy-tailed models
• On scoring maximal ancestral graphs with the max-min hill climbing algorithm
• A survey on causal discovery: theory and practice
• Kernel-based tests for joint independence
• Ensemble framework for causality learning with heterogeneous directed acyclic graphs through the lens of optimization
• Improved baselines for causal structure learning on interventional data
• Structural learning and estimation of joint causal effects among network-dependent variables

Uses Software

• pcalg
• TETRAD

This page was built for publication: Structural intervention distance for evaluating causal graphs

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q5380224)