Causal Explanation of Concept Drift -- A Truly Actionable Approach

TL;DR

This paper introduces a causal explanation framework for concept drift in machine learning, enhancing the actionability of explanations by identifying causally relevant features to enable targeted interventions in changing systems.

Contribution

It extends model-based drift explanations to causal explanations, improving the ability to identify causally relevant features affected by concept drift.

Findings

Framework isolates causally relevant features impacted by drift

Demonstrates practical usefulness across multiple use cases

Enables targeted interventions to prevent model failures

Abstract

In a world that constantly changes, it is crucial to understand how those changes impact different systems, such as industrial manufacturing or critical infrastructure. Explaining critical changes, referred to as concept drift in the field of machine learning, is the first step towards enabling targeted interventions to avoid or correct model failures, as well as malfunctions and errors in the physical world. Therefore, in this work, we extend model-based drift explanations towards causal explanations, which increases the actionability of the provided explanations. We evaluate our explanation strategy on a number of use cases, demonstrating the practical usefulness of our framework, which isolates the causally relevant features impacted by concept drift and, thus, allows for targeted intervention.