Q: 1
Your model has been working fine for the last three months, however recently you notice the model's performance has greatly declined. What seems to have been overlooked in your workflow pipeline?
Options
Discussion
C tbh
Decline like that usually means A. Retraining step was skipped here, not totally sure though since drift is a close distractor.
Don't think it's C here, even though model drift is what happens. A (model retraining) is the step that's actually skipped in these workflows. Easy to mix those up if you go too fast.
A Skipping model retraining is what usually leads to this issue. Drift is the outcome, but retraining's the step that's often missed in real operations. Could be wrong, but that's how I see it.
Its A. Model retraining is the step that keeps your model up to date, so if that's overlooked you'll get declining performance like this. C (model drift) is the result, not what was skipped in the process. Pretty sure about this but let me know if someone sees it differently.
A makes sense if the focus is on what the pipeline missed rather than just labeling the issue. Retraining is that routine checkup step people skip sometimes. Not totally sure though, could maybe argue for D too.
Its A. Model drift (C) is the issue you’d see, but the actual step missed in the workflow is retraining (A). Seen similar questions on practice sets, easy to confuse those two.
Option A Can't ignore retraining, it's the main step missed if performance suddenly tanks after months. Open to other views though.
Doesn't model drift (C) fit better since a performance drop suggests changing input patterns, not just missed retraining?
A Retraining is what you actually skip in these situations, decline like this means the workflow missed that step. If this was asking cause, maybe C, but here pretty sure it's A.
Be respectful. No spam.