Model Training & Adaptation
Feature importance
An estimate of how much a particular piece of information influences an AI's predictions.
Example
An analysis finds that recent purchase history strongly influences a churn score.
Why people use it
It helps teams see which pieces of information strongly influence an AI's predictions.
What you'll hear
“Which information had the biggest influence on the score?”
What this means for you
Ask how importance was calculated before treating it as an explanation.
Can you control it?
No
No direct control. This describes a wider issue, concept or result rather than something you can simply switch on or off in a tool.
Common questions
- Can two checking methods disagree about importance?
- Yes. They may ask different questions about the AI and assign influence differently.
- Can importance change for an individual case?
- Yes. What matters overall may differ from what matters for one particular prediction.
- Can two similar pieces of information confuse the explanation?
- Yes. When they overlap, different methods may divide their apparent importance differently.