Explore the analyst report

IBM is a Leader in The Forrester Wave™: Multimodal Predictive Analytics and Machine Learning, Q3 2020

What is model drift? Why does it matter?

Model drift refers to the degradation of model performance due to changes in data and relationships between input and output variables. It is relatively common for model drift to impact an organization negatively over time or sometimes suddenly. To effectively detect and mitigate drift, organizations can monitor and manage model performance as part of data and AI platform. This integrated approach to data and AI can help you:

  • Track metrics continually and get alerted on drift in accuracy and data consistency.
  • Set targets and track them through development, validation and deployment.
  • Simplify steps to identify business metrics affected by model drift.
  • Minimize the impact of model degradation through automating drift monitoring.

→ Understand drift detection

Dig deeper

Man working on motor

Keep AI models accurate

Get under the hood to learn how the drift monitor works.

Sailboat on the sea

Interactive tour: Track AI models

Experience a hands-on tour of model evaluation and management

Man in greenhouse

Documentation: Validate and monitor models

Get a technical overview of model validation and monitoring.

Take an interactive tour

Explore Watson OpenScale in a self-guided, hands-on demo.

Product image

Detecting drift

Monitor any drift in model accuracy and receive an alert when accuracy degrades past a threshold.

Screenshot of a drift model

Get started with explainable AI

Explore model monitoring and model management. Get model insights and give your business