A Guide to Feature Importance Methods for Scientific Inference

TL;DR

This paper provides a comprehensive review and interpretation guide for feature importance methods in machine learning, aiding scientific inference by clarifying their use, limitations, and future research directions.

Contribution

It offers an extensive review, new proofs, and concrete recommendations for interpreting feature importance methods in scientific research.

Findings

Clarifies different interpretations of FI methods

Provides new proofs for FI interpretation

Recommends best practices for scientific inference

Abstract

While machine learning (ML) models are increasingly used due to their high predictive power, their use in understanding the data-generating process (DGP) is limited. Understanding the DGP requires insights into feature-target associations, which many ML models cannot directly provide due to their opaque internal mechanisms. Feature importance (FI) methods provide useful insights into the DGP under certain conditions. Since the results of different FI methods have different interpretations, selecting the correct FI method for a concrete use case is crucial and still requires expert knowledge. This paper serves as a comprehensive guide to help understand the different interpretations of global FI methods. Through an extensive review of FI methods and providing new proofs regarding their interpretation, we facilitate a thorough understanding of these methods and formulate concrete recommendations for scientific inference. We conclude by discussing options for FI uncertainty estimation and point to directions for future research aiming at full statistical inference from black-box ML models.