Research map / Federated learning and privacy / Privacy attacks and secure learning
On the Gini-impurity Preservation For Privacy Random Forests
XinRan Xie, Man-Jie Yuan, Xuetong Bai, Wei Gao, Zhi-Hua Zhou
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 |
|---|---|---|---|
| encrypt data with Gini-impurity-preserving scheme instead of plaintext | random forest · 20 OpenML datasets | accuracy: win/tie/loss 2/16/2 | No clear effect |
| use Gini-preserving encryption instead of anonymization (AnonyRFs) | random forest · 20 OpenML datasets | accuracy: win/tie/loss 20/0/0; wdbc .9091 -> .9525 | Improved |
| use Gini-preserving encryption instead of differential privacy (DiffPrivRFs) | random forest · 20 OpenML datasets | accuracy: win/tie/loss 20/0/0; wdbc .8998 -> .9525 | Improved |
| use Gini-preserving encryption instead of PPD-ERTs (completely-random splitting) | random forest · 20 OpenML datasets | accuracy: win/tie/loss 14/6/0 | Improved |
| use Gini-preserving encryption instead of PivotRFs (depth-limited) | random forest · 20 OpenML datasets | accuracy: win/tie/loss 10/10/0 | Improved |
| use Gini-preserving encryption instead of MulPRFs | random forest · 20 OpenML datasets | accuracy: win/tie/loss 19/1/0 | No clear effect |
| use Gini-preserving encryption instead of HE-based HEldpRFs | random forest · wdbc, training time | training time: x400 | Improved |
| use Gini-preserving encryption instead of HEldpRFs | random forest · wdbc, prediction time | prediction time: x4, 000 | Improved |
| use Gini-preserving encryption instead of PivotRFs | random forest · wdbc, training time | training time: x10 | Improved |
| use Gini-preserving encryption instead of original RFs (privacy overhead) | random forest · wdbc, training time | training time: x10−3 | Hurt |
| use Gini-preserving encryption instead of order-preserving, DP and anonymization schemes | feature encryption · bitwise leakage matrix | security | Improved |
Closest papers by meaning
- Lumberjack: Better Differentially Private Random Forests through Heavy Hitter Detection in Trees (NeurIPS 2026 · similarity 0.85)
- Trained Random Forests Completely Reveal your Dataset (ICML 2024 · similarity 0.84)
- Improving Private Random Forest Prediction Using Matrix Representation (AAAI 2025 · similarity 0.84)
- Privacy-Preserving Embedding via Look-up Table Evaluation with Fully Homomorphic Encryption (ICML 2024 · similarity 0.81)
- Provable Privacy with Non-Private Pre-Processing (ICML 2024 · similarity 0.81)
- Purifying Approximate Differential Privacy with Randomized Post-processing (NeurIPS 2025 · similarity 0.80)
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.
