Welcome to the workshop!
Workshop
In ProgressWelcome to Detecting and responding to model drift in production
Today's icebreaker:
When the model drifts…
Imagine your shopping app drifts and starts making bizarre recommendations (for example, 50 cans of dog food when you don’t own a dog).
What other funny or weird recommendations might a ‘drifting’ shopping app give you?
Type your ideas in the chat.
Today's agenda:
✨
Review:
Recap key concepts.
✨
Practical exercise:
ShopSmart drift response plan.
✨
Closing:
Wrap-up and reflection.
Today's learning objectives:
- Detect and classify different types of model drift in production ML systems.
- Design monitoring dashboards that link drift indicators to performance metrics and business outcomes.
- Propose maintenance and testing strategies to respond to drift and ensure safe deployment of model updates.
Workshop slides
Preview or download a copy of the Workshop 1 slides:
PDF preview loads in your browser…
If it doesn’t appear, download it here.
In progress