Causal Discovery for Explainable AI: A Dual-Encoding Approach

TL;DR

This paper introduces a dual-encoding causal discovery method that improves the identification of causal relationships in explainable AI, especially with categorical data, by combining multiple encoding strategies and majority voting.

Contribution

It presents a novel dual-encoding approach that enhances causal discovery accuracy for categorical variables in explainable AI.

Findings

Successfully applied to Titanic dataset

Identifies causal structures consistent with established methods

Addresses numerical instability in causal testing

Abstract

Understanding causal relationships among features is fundamental for explaining machine learning model decisions. However, traditional causal discovery methods face challenges with categorical variables due to numerical instability in conditional independence testing. We propose a dual-encoding causal discovery approach that addresses these limitations by running constraint-based algorithms with complementary encoding strategies and merging results through majority voting. Applied to the Titanic dataset, our method identifies causal structures that align with established explainable methods.