Entropic Causal Inference: Identifiability and Finite Sample Results

TL;DR

This paper proves that causal direction between two categorical variables can be identified under certain entropy conditions of the exogenous variable, providing finite sample guarantees and practical robustness analyses.

Contribution

It establishes theoretical identifiability results for entropic causal inference and demonstrates finite sample guarantees for the algorithmic approach.

Findings

Causal direction is identifiable when exogenous entropy does not scale with observed variables.

Finite sample guarantees are provided for the causal inference algorithm.

The method is robust to relaxing assumptions and performs well on real datasets.

Abstract

Entropic causal inference is a framework for inferring the causal direction between two categorical variables from observational data. The central assumption is that the amount of unobserved randomness in the system is not too large. This unobserved randomness is measured by the entropy of the exogenous variable in the underlying structural causal model, which governs the causal relation between the observed variables. Kocaoglu et al. conjectured that the causal direction is identifiable when the entropy of the exogenous variable is not too large. In this paper, we prove a variant of their conjecture. Namely, we show that for almost all causal models where the exogenous variable has entropy that does not scale with the number of states of the observed variables, the causal direction is identifiable from observational data. We also consider the minimum entropy coupling-based algorithmic approach presented by Kocaoglu et al., and for the first time demonstrate algorithmic identifiability guarantees using a finite number of samples. We conduct extensive experiments to evaluate the robustness of the method to relaxing some of the assumptions in our theory and demonstrate that both the constant-entropy exogenous variable and the no latent confounder assumptions can be relaxed in practice. We also empirically characterize the number of observational samples needed for causal identification. Finally, we apply the algorithm on Tuebingen cause-effect pairs dataset.