In SAS Model Manager, which term describes the automated process of detecting when a model's predictive performance has degraded enough to trigger an alert or retraining workflow?
Select an answer to reveal the explanation.
Short Explanation
You want model drift monitoring because it automatically tracks performance metrics over time and flags statistically significant degradation, such as concept drift or data drift, that can trigger alerts or retraining. Model validation is a one-time evaluation step before deployment, not the ongoing detection process.
Full Explanation
Model drift monitoring is correct because it is the automated process of tracking performance metrics over time and detecting statistically significant degradation, including concept drift or data drift, that triggers alerts or retraining workflows in SAS Model Manager. Model validation is wrong because it is a one-time evaluation step before deployment rather than continuous monitoring. Model versioning is wrong because it tracks different iterations of a model in the registry, not performance degradation. Model packaging is wrong because it bundles score code and resources for deployment, not automated detection of degraded predictive performance.