Model Drift
Model drift is the decline in an AI model's performance over time because real-world data or relationships change from those it was trained on.
What is Model Drift?
Data drift is a change in input data; concept drift is a change in the relationship between inputs and outcomes. Continuous monitoring is needed to detect it.
How it works
Teams track input statistics and accuracy over time and retrain models when thresholds are breached.
← All Artificial Intelligence & Superintelligence definitions · ← Back to the A–Z definitions