Causal Shapley Values: Exploiting Causal Knowledge to Explain Individual Predictions of Complex Models

TL;DR

This paper introduces causal Shapley values that leverage causal knowledge and Pearl's do-calculus to improve feature attribution explanations for complex models, especially when feature independence assumptions are violated.

Contribution

It proposes a novel framework for computing Shapley values using causal graphs and do-calculus, enabling separation of direct and indirect feature effects.

Findings

Causal Shapley values generalize existing methods without losing desirable properties.

The framework can handle partial causal information.

Application to real-world data demonstrates practical utility.

Abstract

Shapley values underlie one of the most popular model-agnostic methods within explainable artificial intelligence. These values are designed to attribute the difference between a model's prediction and an average baseline to the different features used as input to the model. Being based on solid game-theoretic principles, Shapley values uniquely satisfy several desirable properties, which is why they are increasingly used to explain the predictions of possibly complex and highly non-linear machine learning models. Shapley values are well calibrated to a user's intuition when features are independent, but may lead to undesirable, counterintuitive explanations when the independence assumption is violated. In this paper, we propose a novel framework for computing Shapley values that generalizes recent work that aims to circumvent the independence assumption. By employing Pearl's do-calculus, we show how these 'causal' Shapley values can be derived for general causal graphs without sacrificing any of their desirable properties. Moreover, causal Shapley values enable us to separate the contribution of direct and indirect effects. We provide a practical implementation for computing causal Shapley values based on causal chain graphs when only partial information is available and illustrate their utility on a real-world example.