Explaining the root causes of unit-level changes

TL;DR

This paper introduces two counterfactual-based methods utilizing Shapley values to explain the causes of unit-level changes in output variables, addressing a gap in existing explainable AI techniques.

Contribution

It proposes novel methods for explaining change in individual outputs by combining counterfactuals and Shapley values, satisfying key axioms for attribution methods.

Findings

Methods are reliable and scalable based on simulations.

Case study successfully identifies drivers of earnings change.

Approach enhances interpretability of unit-level change explanations.

Abstract

Existing methods of explainable AI and interpretable ML cannot explain change in the values of an output variable for a statistical unit in terms of the change in the input values and the change in the "mechanism" (the function transforming input to output). We propose two methods based on counterfactuals for explaining unit-level changes at various input granularities using the concept of Shapley values from game theory. These methods satisfy two key axioms desirable for any unit-level change attribution method. Through simulations, we study the reliability and the scalability of the proposed methods. We get sensible results from a case study on identifying the drivers of the change in the earnings for individuals in the US.