What is Model Drift?

MLOps

Model Drift — When a deployed model performance degrades over time because the real-world data distribution changes from what the model was trained on. Requires monitoring and retraining.

FAQ

What is model drift?

When your model gets worse over time because real-world data changes. Example: a fraud model trained on 2024 patterns failing on 2026 fraud tactics.

How to detect drift?

Monitor prediction distributions, accuracy metrics, and input feature distributions. Tools: Evidently AI, WhyLabs, custom dashboards.

Related Terms

Learn Model Drift in depth

Free hands-on course with code examples and Google Colab notebooks.

Start Course →