Skip to contents

Auto XAI: Automatic Explainability for Biomarker Models

Source: R/AutoXAI.R
AutoXAI.Rd

Provides automatic interpretability outputs for mlr3 models using DALEX. Designed for high-dimensional omics data with correlation-aware warnings.

Details

Key features: - mlr3 to DALEX adapter for probabilistic classifiers - Permutation-based feature importance - Partial Dependence Plots (PDPs) for top features - SHAP values for individual predictions - Correlation diagnostics and warnings - Automated report generation