PROBLEM: Alert triage mixes two different failures — missing real signals and over-firing on noise — but no standard record separates them, so a triage rule that is insensitive and one that is merely biased cannot be told apart.
SOLUTION: A signal-detection calibration record that logs each triage rule with its sensitivity and its response criterion measured separately, so accuracy and bias are each checkable.
EVIDENCE BASIS: DOCUMENTED — signal detection theory (Green & Swets, 1966): detection separates into sensitivity (d′) and response criterion (F3).
| Idea priority | Low |
| Needed By | Not sure -- Just thought it was cool |
By clicking the "Post Comment" or "Submit Idea" button, you are agreeing to the IBM Ideas Portal Terms of Use.
Do not place IBM confidential, company confidential, or personal information into any field.