Interview Guides
You are monitoring a manufacturing signal from an inline sensor and need to decide whether unusual readings reflect normal measurement noise or a real process change. You want a statistically defensible way to separate transient noise from a true anomaly.
What statistical methods would you use to differentiate between sensor noise and a true manufacturing anomaly?
