AI Paper Atlas

Research map / Probabilistic and scientific modeling / Treatment effects

Stable Estimation of Heterogeneous Treatment Effects

Anpeng Wu, Kun Kuang, Ruoxuan Xiong, Bo Li, Fei Wu

ICML 2023 · Probabilistic and scientific modeling › Treatment effects · approach: Meta-learners and partial identification

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
Uniform resampling + epsilon-greedy matching (StableCFR) vs. no imbalance technique (VANILLA)two-head network on Syn-0.8 · Syn-0.8 OOD dataPEHE: 0.500 -> 0.299Improved
Uniform resampling + epsilon-greedy matching (StableCFR)two-head network on Syn-0.8 · Syn-0.8 ID dataPEHE: 0.089 -> 0.099Hurt
Uniform resampling + epsilon-greedy matching (StableCFR)two-head network on Semi-PM-CMR · Semi-PM-CMR OOD dataPEHE: 0.548 (IPWNet) -> 0.489; about 11% reduction vs SOTAImproved
Uniform resampling + epsilon-greedy matching (StableCFR)two-head network on Semi-PM-CMR · Semi-PM-CMR ID dataPEHE: 0.131 (VANILLA) vs 0.139No clear effect
Uniform resampling + epsilon-greedy matching (StableCFR)two-head network on Syn-0.8 · Syn-0.8 OOD dataMAXE: 7.628 (DeR-CFR) -> 5.889Improved
Pure uniform sampling with nearest neighbor (epsilon=1) vs. with epsilon-greedy matching (ablation)StableCFR · Syn-0.8 (ablation of matching)qualitative PEHE and varianceImproved
Conventional CFR-based balancing methods (CFRNet, SITE, CFR-ISW) vs. StableCFR without regularityrepresentation balancing baselines · Syn-0.8 OOD dataPEHE: CFRNet 5.990, SITE 7.008 vs StableCFR 0.299Hurt
Imbalance regression reweighting/balanced MSE (Reweight, BMC, GAI, DIRNet) vs. VANILLAregression baselines on Syn-0.8 · Syn-0.8 OOD dataPEHE: Reweight 1.205, BMC 1.574, GAI 2.136, DIRNet 0.860 vs VANILLA 0.500Hurt

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