AI Paper Atlas

Research map / Learning theory and algorithms / Prediction and decision losses

A Unified Framework for Random Forest Prediction Error Estimation

Benjamin Lu, Johanna Hardin

JMLR 2021 · Learning theory and algorithms › Prediction and decision losses · approach: Probabilistic calibration · 22 citations (OpenAlex)

Read the paper

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
replace boosting bias correction with the out-of-bag weighted bias correctionrandom forest · ServoMSPE: 11.779 -> 17.601Hurt
apply weighted OOB bias correctionrandom forest · Step (synthetic)MSB: 0.814 -> 0.222Improved
apply weighted OOB bias correctionrandom forest · Friedman (synthetic)MSPE: 7.018 -> 4.927Improved
apply weighted OOB bias correctionrandom forest · BostonMSPE: 8.001 -> 6.973Improved
apply weighted OOB bias correctionrandom forest · Baseline (synthetic)MSPE: 1.063 -> 1.085No clear effect
increase response variance to 100 (noisy setting)bias-corrected RF vs boosting · Friedman (noised)MSPE: 119.532 -> 113.898Improved
weight OOB errors by cohabitation instead of equal weightsprediction intervals · Clusteredinterval width: 15.83 -> 13.94Improved
use OOB error quantiles instead of response quantiles (QRF)prediction intervals · Friedmaninterval width: 36.27 -> 22.01Improved
use stringent sample-split estimator with three times the training data (ablation)stringent bias-corrected RF · Friedman (synthetic)MSPE: 7.327 -> 4.446Improved
use stringent sample-split estimator with three times the training data (ablation)stringent PI estimator · Clusteredinterval width: 16.57 -> 13.87Improved
adapt the method to generalized random forest treesPI estimator · Clusteredinterval width: 41.27 -> 11.94Improved
adapt the method to generalized random forest treesPI estimator · Bostoncoverage: 0.994 -> 0.965Improved

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