Consistency of Causal Inference under the Additive Noise Model

TL;DR

This paper investigates the statistical consistency of causal inference methods based on the Additive Noise Model, providing conditions under which these methods reliably determine causal direction in nonparametric settings.

Contribution

It establishes general conditions ensuring the consistency of additive noise model-based causal inference methods, addressing a gap in theoretical understanding.

Findings

Derives conditions for consistent causal inference under the additive noise model.

Provides theoretical guarantees for nonparametric causal inference methods.

Enhances understanding of when additive noise models yield reliable causal directions.

Abstract

We analyze a family of methods for statistical causal inference from sample under the so-called Additive Noise Model. While most work on the subject has concentrated on establishing the soundness of the Additive Noise Model, the statistical consistency of the resulting inference methods has received little attention. We derive general conditions under which the given family of inference methods consistently infers the causal direction in a nonparametric setting.