Research map / Learning theory and algorithms / Prediction and decision losses
Hybrid Predictive Models: When an Interpretable Model Collaborates with a Black-box Model
Tong Wang, Qihang Lin
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 |
|---|---|---|---|
| hybrid rule set replacing black-box on 71.6% of data | Adaboost · Census | accuracy: 85.0 | Improved |
| send 80% of data to interpretable model | XGBoost · Juvenile | accuracy: 90.6 | No clear effect |
| route data to interpretable rules or linear model up to about 50% transparency | black-box models · Census | accuracy | No clear effect |
| route more than 80% of data to interpretable model | black-box models · Juvenile | accuracy | Hurt |
| hybrid rule sets at low transparency | black-box models · Movie success prediction | accuracy | Improved |
| use hybrid linear model instead of hybrid rule set for training time | hybrid model · Census | training time: 10m 55s ± 45s -> 1m 5s ± 1s | Improved |
| use hybrid linear model instead of hybrid rule set for training time | hybrid model · Stop-and-Frisk | training time: 41m 50s ± 5s -> 2m 32s ± 3s | Improved |
| use Lasso instead of black-box models | interpretable baseline · Juvenile | accuracy: 0.904 -> 0.909 | Improved |
| use XGBoost instead of RF | black-box · Coupon | accuracy: 0.755 -> 0.703 | Hurt |
Closest papers by meaning
- From Black-box to Causal-box: Towards Building More Interpretable Models (NeurIPS 2025 · similarity 0.85)
- Beyond Concept Bottleneck Models: How to Make Black Boxes Intervenable? (NeurIPS 2024 · similarity 0.84)
- Enhancing Performance of Explainable AI Models with Constrained Concept Refinement (ICML 2025 · similarity 0.84)
- Explainable Neural Networks with Guarantee: A Sparse Estimation Approach (AAAI 2025 · similarity 0.83)
- Multistage Defer Trees for Hybrid Interpretability: If at First You Can't Succeed, Tree Again (NeurIPS 2026 · similarity 0.83)
- Local vs. Global Interpretability: A Computational Complexity Perspective (ICML 2024 · 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.
