Research map / Adversarial robustness and security / Certified and empirical robustness
Robust Decision Trees Against Adversarial Examples
Read the paperCheck your idea against 108,967 papers
AI-extracted summary. Written by AI from the paper's title only (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 worst-case robust split scoring with ℓ∞ perturbation radius ε | XGBoost GBDT · MNIST | average ℓ∞ distortion (Cheng): .072 -> .373 | Helped |
| Use worst-case robust split scoring with ℓ∞ perturbation radius ε | XGBoost GBDT · Fashion-MNIST | average ℓ∞ distortion (Cheng): .049 -> .156 | Helped |
| Use worst-case robust split scoring with ℓ∞ perturbation radius ε | XGBoost GBDT · HIGGS | average ℓ∞ distortion (Cheng): .014 -> .022 | Helped |
| Use worst-case robust split scoring with ℓ∞ perturbation radius ε | XGBoost GBDT · MNIST | test accuracy: .980 -> .980 | No clear effect |
| Use worst-case robust split scoring with ℓ∞ perturbation radius ε | XGBoost GBDT · HIGGS | test accuracy: .760 -> .709 | Hurt |
| Use worst-case robust split scoring with ℓ∞ perturbation radius ε | information-gain decision tree · ionosphere | test accuracy: .929 -> .986 | Helped |
| Use worst-case robust split scoring with ℓ∞ perturbation radius ε | information-gain decision tree · diabetes | test accuracy: .747 -> .688 | Hurt |
| Reduce tree depth instead of using robust split training (ablation) | naturally trained GBDT · MNIST | adversarial robustness | No clear effect |
| Add robust split training from the first boosting tree (ablation) | MNIST GBDT · MNIST | ℓ∞ distortion: .033 -> .330 | Helped |
Closest papers by meaning
- Faster Repeated Evasion Attacks in Tree Ensembles (NeurIPS 2024 · similarity 0.88)
- TTTS: Tree Test Time Simulation for Enhancing Decision Tree Robustness against Adversarial Examples (AAAI 2024 · similarity 0.88)
- Coevolutionary Algorithm for Building Robust Decision Trees under Minimax Regret (AAAI 2024 · similarity 0.87)
- Robustness Verification of Multi-Class Tree Ensembles (AAAI 2024 · similarity 0.86)
- Fully-Dynamic Approximate Decision Trees With Worst-Case Update Time Guarantees (ICML 2024 · similarity 0.85)
- Learning Decision Trees and Forests with Algorithmic Recourse (ICML 2024 · similarity 0.85)
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.