AI Paper Atlas

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

JMLR 2021 · Learning theory and algorithms › Prediction and decision losses · approach: Provable post-hoc explanations · 26 citations (OpenAlex)

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 changedWhereMeasuredEffect
hybrid rule set replacing black-box on 71.6% of dataAdaboost · Censusaccuracy: 85.0Improved
send 80% of data to interpretable modelXGBoost · Juvenileaccuracy: 90.6No clear effect
route data to interpretable rules or linear model up to about 50% transparencyblack-box models · CensusaccuracyNo clear effect
route more than 80% of data to interpretable modelblack-box models · JuvenileaccuracyHurt
hybrid rule sets at low transparencyblack-box models · Movie success predictionaccuracyImproved
use hybrid linear model instead of hybrid rule set for training timehybrid model · Censustraining time: 10m 55s ± 45s -> 1m 5s ± 1sImproved
use hybrid linear model instead of hybrid rule set for training timehybrid model · Stop-and-Frisktraining time: 41m 50s ± 5s -> 2m 32s ± 3sImproved
use Lasso instead of black-box modelsinterpretable baseline · Juvenileaccuracy: 0.904 -> 0.909Improved
use XGBoost instead of RFblack-box · Couponaccuracy: 0.755 -> 0.703Hurt

Closest papers by meaning

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.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy