Model Training & Adaptation
Logistic regression
A method often used to estimate how likely something is to belong to a particular group.
Example
A system estimates an email's chance of being spam.
Why people use it
It provides a relatively simple way to estimate how likely a category is.
What you'll hear
“How likely is this message to be spam?”
What this means for you
Check what score causes a message to be blocked and how often that choice is wrong.
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
- Why is regression in the name?
- The method estimates a numerical chance, which can then be used to choose a group.
- Does a score of 80% guarantee the outcome?
- No. It describes an estimate, not a promise about that individual case.
- Can it work with more than two categories?
- Related versions can handle several categories, although the familiar basic form handles two.