AI Paper Atlas

Research map / Probabilistic and scientific modeling / Treatment effects

A Relative Error-Based Evaluation Framework of Heterogeneous Treatment Effect Estimators

Jiayi Guo, Haoxuan Li, YE TIAN, Peng Wu

ICLR 2026 · Probabilistic and scientific modeling › Treatment effects · approach: Meta-learners and partial identification · 1 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
use proposed network with new loss and balance regularizers vs best baselineHTE estimation · IHDP in-samplesqrt(PEHE): 0.741 -> 0.638Improved
proposed method vs best baselineHTE estimation · Twins in-samplesqrt(PEHE): 0.290 -> 0.284Improved
proposed method vs LinDML on Twins ATEHTE estimation · Twins in-sampleATE error: 0.013 -> 0.009Improved
use proposed nuisance network instead of linear regressionrelative error estimation · IHDPselection accuracy: 0.44 -> 0.80Improved
use proposed nuisance network instead of gradient boostingrelative error estimation · Twinsselection accuracy: 0.86 -> 0.94Improved
conventional nuisance estimatorsrelative error estimation · IHDP/Twinscoverage: 0.94 vs 0.96No clear effect
add constraint loss L_const (ablation)proposed network · IHDP in-samplesqrt(PEHE): 3.495 -> 0.638Improved
add cross-entropy loss L_ce (ablation)proposed network · IHDP in-samplesqrt(PEHE): 0.725 -> 0.638Improved
add L_ce on Twins (ablation)proposed network · Twins in-samplesqrt(PEHE): 0.284 -> 0.284No clear effect
increase lambda_2 from 0.01 to 1 (ablation)proposed network · IHDP selectionselection accuracy: 0.50 -> 0.80Improved
vary lambda_2 between 0.5 and 5 (ablation)proposed network · IHDPsqrt(PEHE) in: 0.714 / 0.638 / 0.715No clear effect

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