Knowledge Maxxing← Back to topic library
ML MaxxingIntermediate75 sec
Feature importance
Feature importance measures how much each input feature contributes to a model's predictions overall, ranking features by their typical influence.
Prerequisite: Feature
What matters
Understanding which features drive a model's decisions helps catch problems like reliance on a leaked or spurious feature, and helps explain the model's behavior to others.
Example
A feature-importance ranking might reveal that a loan model relies heavily on income and barely at all on requested loan amount.
Common misconception
High feature importance does not prove a feature causes the outcome; it only shows the model relies on it statistically, which can reflect correlation rather than causation.
Recall question
Why doesn't high feature importance prove that a feature causes the predicted outcome?
Keywords and related concepts
feature importancemodel explanationinfluenceshap value