Research map / Probabilistic and scientific modeling / Molecules and proteins
Exact Functional ANOVA Decomposition for Categorical Inputs Models
Baptiste Ferrere, Nicolas Bousquet, Fabrice Gamboa, Jean-Michel Loubes, Joseph Muré
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 |
|---|---|---|---|
| raise rank from main effects to r_high (ablation) | closed-form ANOVA · Poker Hand | R2: 0.003 -> 0.79 | Improved |
| raise rank from main effects to r_high (ablation) | closed-form ANOVA · Connect-4 | R2: 0.45 -> 0.70 | Improved |
| raise rank from main effects to r_high (ablation) | closed-form ANOVA · Dota2 | R2: 0.36 -> 0.41 | Improved |
| raise rank from main effects to r_high (ablation) | closed-form ANOVA · Dota2 | time: 23s -> 39min | Hurt |
| raise rank from main effects to r_high (ablation) | closed-form ANOVA · Poker Hand | time: 1.8s -> 10min | Hurt |
| raise rank from 674 to 2500 (ablation) | closed-form ANOVA · Binarized MNIST | R2: 0.54 -> 0.77 | Improved |
| raise rank from 2500 to 5000 (ablation) | closed-form ANOVA · Binarized MNIST | R2: 0.77 -> 0.83 | Improved |
| raise rank from 5000 to 10000 (ablation) | closed-form ANOVA · Binarized MNIST | R2: 0.83 -> 0.86 | Improved |
| raise rank from 5000 to 10000 (ablation) | closed-form ANOVA · Binarized MNIST | relative MSE: 15 % -> 12 % | Improved |
| use closed-form ANOVA instead of TreeHFD | orthogonality · XGBoost surrogate d=2 | max inner product: 1.77 x 10^-4 -> 1.54 x 10^-19 | Improved |
| use closed-form ANOVA instead of TreeHFD | orthogonality · XGBoost surrogate d=10 | max inner product: 6.65 x 10^-4 -> 2.72 x 10^-16 | Improved |
| replace KernelSHAP by exact ANOVA Shapley | Shapley values · Car Evaluation and Nursery | ISE: 0.00 to 4.4 (100 x ISE) | No clear effect |
Closest papers by meaning
- Tree Ensemble Explainability through the Hoeffding Functional Decomposition and TreeHFD Algorithm (NeurIPS 2025 · similarity 0.84)
- A Unified View of FANOVA: A Comprehensive Bayesian Framework for Component Selection and Estimation (ICML 2024 · similarity 0.84)
- Fast Estimation of Partial Dependence Functions using Trees (ICML 2025 · similarity 0.83)
- Practical do-Shapley Explanations with Estimand-Agnostic Causal Inference (NeurIPS 2025 · similarity 0.82)
- Tensor Product Neural Networks for Functional ANOVA Model (ICML 2025 · similarity 0.82)
- Regional Explanations: Bridging Local and Global Variable Importance (NeurIPS 2025 · similarity 0.82)
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.
