Trustworthy Feature Importance Avoids Unrestricted Permutations

arXiv:2604.11253v1 Announce Type: new
Abstract: Feature importance methods using unrestricted permutations are flawed due to extrapolation errors; such errors appear in all non-trivial variable importance approaches. We propose three new approaches: conditional model reliance and Knockoffs with Gaussian transformation, and restricted ALE plot designs. Theoretical and numerical results show our strategies reduce/eliminate extrapolation.