Statistical vs. computational monitoring
It is best practice to monitor the machine learning model as soon as it is deployed. The performance of machine learning models can be monitored statistically as well as computationally. Monitoring in different ways is important because your model can be live in production and seem fine, while the predictions can be wrong.
In this exercise, you will classify which metric belongs to which type of performance monitoring.
This exercise is part of the course
MLOps Concepts
Hands-on interactive exercise
Turn theory into action with one of our interactive exercises
