Research map / Probabilistic and scientific modeling / Molecules and proteins
Statistically Valid Variable Importance Assessment through Conditional Permutations
Ahmad Chamma, Denis Engemann, Bertrand Thirion
Read the paperCheck your idea against 137,217 papers
AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| Use conditional permutation instead of standard permutation | DNN variable importance · Simulations with increasing variable correlation | type-I error control | Improved |
| Use conditional permutation instead of standard permutation | DNN variable importance · Simulations with increasing variable correlation | AUC | No clear effect |
| Use conditional permutation instead of standard permutation | DNN variable importance · Five simulated data-generating scenarios | type-I error | Improved |
| Use conditional permutation instead of standard permutation | DNN variable importance · Five simulated data-generating scenarios | AUC | Improved |
| Use conditional permutation instead of standard permutation | DNN variable importance · UK Biobank-based simulation | AUC | Improved |
| Use conditional permutation instead of standard permutation | DNN variable importance · UK Biobank-based simulation | power | Improved |
| Use conditional permutation instead of standard permutation | DNN variable importance · UK Biobank-based simulation | type-I error | No clear effect |
| Use conditional permutation instead of standard permutation | DNN variable importance · Large-scale simulation, correlation coefficient 0.8 | type-I error | Improved |
| Use conditional permutation instead of standard permutation | DNN variable importance · UK Biobank age prediction | selected brain inputs: 16 out of 66 | Mixed |
| Use conditional permutation instead of standard permutation | DNN variable importance · Simulations across data-generating scenarios | power | No clear effect |
| Use conditional permutation instead of standard permutation | DNN variable importance · High-correlation simulations | computation time | Hurt |
| Use conditional permutation with an RF conditional model rather than DNN as the importance learner | CPI variable importance · Simulated scenarios | AUC | Hurt |
Closest papers by meaning
- Variable Importance in High-Dimensional Settings Requires Grouping (AAAI 2024 · similarity 0.93)
- Measuring Variable Importance in Heterogeneous Treatment Effects with Confidence (ICML 2025 · similarity 0.88)
- Covered Information Disentanglement: Model Transparency via Unbiased Permutation Importance (AAAI 2022 · similarity 0.86)
- All Models are Wrong, but Many are Useful: Learning a Variable's Importance by Studying an Entire Class of Prediction Models Simultaneously (JMLR 2019 · similarity 0.86)
- Total Variation Floodgate for Variable Importance Inference in Classification (ICML 2024 · similarity 0.84)
- Permutation-based Rank Test in the Presence of Discretization and Application in Causal Discovery with Mixed Data (ICML 2025 · similarity 0.83)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.
