Machine Learning Model Drift Detection Via Weak Data Slices

TL;DR

This paper introduces a label-efficient method for detecting machine learning model drift by analyzing feature space rules, enabling early identification of performance degradation without relying on costly labels.

Contribution

It proposes a novel drift detection approach using data slices based on feature space rules, addressing the challenge of label scarcity in model monitoring.

Findings

Method effectively predicts model performance changes

Detects data shifts without requiring labels

Shows promising experimental results

Abstract

Detecting drift in performance of Machine Learning (ML) models is an acknowledged challenge. For ML models to become an integral part of business applications it is essential to detect when an ML model drifts away from acceptable operation. However, it is often the case that actual labels are difficult and expensive to get, for example, because they require expert judgment. Therefore, there is a need for methods that detect likely degradation in ML operation without labels. We propose a method that utilizes feature space rules, called data slices, for drift detection. We provide experimental indications that our method is likely to identify that the ML model will likely change in performance, based on changes in the underlying data.