Feature Importance Methods for Global Explanations
Conditional Feature Importance
Check group specification
ARF-based Conditional Sampler
(experimental) Conditional Inference Tree Conditional Sampler
Gaussian Conditional Sampler
k-Nearest Neighbors Conditional Sampler
Conditional SAGE
Conditional Feature Sampler
Feature Importance Method Class
Feature Sampler Class
Gaussian Knockoff Conditional Sampler
Knockoff Sampler
Leave-One-Covariate-Out (LOCO)
Marginal Permutation Sampler
Marginal Reference Sampler
Marginal SAGE
Marginal Sampler Base Class
Default value for NULL
Perturbation Feature Importance Base Class
Permutation Feature Importance
Print an Rd-formatted bib entry
Relative Feature Importance
Aggregate Predictions by Coalition and Test Instance
Batch Predict for SAGE
Shapley Additive Global Importance (SAGE) Base Class
Simulate data as in Ewald et al. (2024)
Simulation DGPs for Feature Importance Method Comparison
Create Feature Selection Design Matrix
Williamson's Variable Importance Measure (WVIM)
xplainfi Package Options
xplainfi: Feature Importance Methods for Global Explanations
Provides a consistent interface for common feature importance methods as described in Ewald et al. (2024) <doi:10.1007/978-3-031-63797-1_22>, including permutation feature importance (PFI), conditional and relative feature importance (CFI, RFI), leave one covariate out (LOCO), and Shapley additive global importance (SAGE), as well as feature sampling mechanisms to support conditional importance methods.
Useful links