Machine Learning Model Drift Detection Via Weak Data Slices
This addresses the challenge of maintaining ML models in business applications where labels are scarce, though it appears incremental as it builds on existing drift detection concepts.
The paper tackles the problem of detecting performance drift in machine learning models without requiring labels, which are often expensive to obtain, by proposing a method that uses feature space rules called data slices, and provides experimental indications that it can identify likely performance changes based on data shifts.
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.