Model Training & Adaptation
Binary classification
Sorting something into one of two possible groups.
Example
A system decides whether an email is spam or ordinary mail.
Why people use it
It supports decisions that need one of two possible labels.
What you'll hear
“Is this message spam or not?”
What this means for you
Consider the cost of blocking real messages as well as letting unwanted ones through.
Can you control it?
Developer-only
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 a binary decision mean the AI system is certain?
- No. An uncertain score may still be converted into a yes-or-no label.
- Must the two groups be equally common?
- No. One group can be much rarer, which may make mistakes easier to overlook.
- Can a score sit close to the dividing line?
- Yes. Two nearly identical scores can receive different labels if they fall on opposite sides.