Glossary term
Glossary term
Foundations
A type of variable importance that evaluates the increase in the prediction error of a model after permuting the feature's values. Permutation variable importance is a model-independent metric.
Created for this library
A risk modeling team uses permutation variable importances to communicate the drivers of its boosted-tree scorecard to business reviewers.
A retail forecasting team uses permutation variable importances to confirm that key business drivers remain influential after retraining.
A churn team reports permutation variable importances each release so the retention team can adjust outreach playbooks to the latest drivers.
Definition source: Google for Developers Machine Learning Glossary | Creative Commons Attribution 4.0 License