Model monitoring
Watching a live model’s inputs, outputs and performance.
Monitoring tracks accuracy, drift, fairness measures, error rates and unusual outputs against thresholds, with alerts, owners and documented responses.
On the exam
A monitoring dashboard nobody acts on is a weak control. Look for thresholds, ownership and evidence of action.
Related terms
- Data drift and model driftWhen a model gets worse because the world changed, not the code.
- Bias and fairnessWhether an AI system treats groups of people unfairly.
Test yourself with today’s question of the day.