1 / 5
The data science team will ___ the model on the updated dataset over the weekend.
-
-
-
-
You train a model. Train is the precise ML collocation; the others are informal or incorrect. So train a model.
2 / 5
After training, we use a held-out test set to ___ the model's performance.
-
-
-
-
You evaluate performance. Evaluate is the standard ML collocation; the others are informal or less precise. So evaluate performance.
3 / 5
Once the model passes benchmarks, the MLOps team will ___ it to production.
-
-
-
-
You deploy a model to production. Deploy is the precise ML-engineering collocation; the others are informal or less specific. So deploy to production.
4 / 5
The team set up a pipeline to automatically ___ drift in the model's predictions over time.
-
-
-
-
You monitor drift. Monitor is the standard MLOps collocation; the others are informal or incorrect. So monitor drift.
5 / 5
When accuracy drops below threshold, the system triggers a job to ___ the model on new data.
-
-
-
-
You retrain a model on new data. Retrain is the precise ML collocation; the others are informal or incorrect. So retrain on new data.