Ultra-marginal Feature Importance: Learning from Data with Causal Guarantees

TL;DR

This paper introduces ultra-marginal feature importance (UMFI), a new method that improves causal interpretability and computational efficiency over existing approaches by leveraging dependence removal techniques.

Contribution

The paper develops UMFI, enhancing theoretical properties, performance, and runtime of marginal contribution feature importance by integrating dependence removal techniques from AI fairness.

Findings

UMFI satisfies proposed axioms for causal and associative explanations.

UMFI outperforms MCI on real and simulated data, especially with correlated features.

UMFI reduces runtime from exponential to super-linear.

Abstract

Scientists frequently prioritize learning from data rather than training the best possible model; however, research in machine learning often prioritizes the latter. Marginal contribution feature importance (MCI) was developed to break this trend by providing a useful framework for quantifying the relationships in data. In this work, we aim to improve upon the theoretical properties, performance, and runtime of MCI by introducing ultra-marginal feature importance (UMFI), which uses dependence removal techniques from the AI fairness literature as its foundation. We first propose axioms for feature importance methods that seek to explain the causal and associative relationships in data, and we prove that UMFI satisfies these axioms under basic assumptions. We then show on real and simulated data that UMFI performs better than MCI, especially in the presence of correlated interactions and unrelated features, while partially learning the structure of the causal graph and reducing the exponential runtime of MCI to super-linear.