Evaluation & Quality
F1 score
A metric that balances precision and recall in a single number.
Example
A team uses F1 when both missed fraud and false alarms matter.
Why it matters
It makes competing detection systems easier to compare with a single headline measure.
You'll hear it
“The combined score improved, but which kind of mistake changed?”
Can you control it?
Set by the team
The people building or running the AI choose this setup. An everyday user generally needs their help to change how this part works.
Common questions
- Does F1 give false alarms and misses any real-world cost?
- No. The score does not automatically reflect which mistakes are more harmful in your situation.
- Can two systems share an F1 score but make different mistakes?
- Yes. Different balances of finding cases and avoiding false alarms can produce the same number.
- Does F1 include correct all-clear decisions?
- Not directly. It focuses on detections, missed cases and false alarms rather than every possible outcome.