Model output distribution shifts. What are root causes and next steps?
This tests covariate vs label shift vs concept drift when outputs shift. A strong answer checks features before labels, then feedback loops or staleness. A red flag is generic drift without separating P(X), P(Y), and P(Y|X).
This tests if you can separate covariate shift, label shift, and concept drift when outputs shift, and if you can triage upstream feature changes vs model staleness. A strong answer first checks input feature distributions, then label priors, then concept drift or degenerate feedback loops. It also mentions checking slice-level metrics and pipeline bugs. A red flag is generic drift without distinguishing P(X), P(Y), or P(Y|X), or retraining before root cause analysis.
Read the original → huyenchip.com
- #mlops
- #data-drift
- #monitoring
- #concept-drift
- #covariate-shift
Get five bites like this every day.
Tezvyn delivers a daily feed of 60-second tech bites with quizzes to lock in what you learn.