AI Paper Atlas

Research map / Probabilistic and scientific modeling / Treatment effects

Generalized Optimal Matching Methods for Causal Inference

Nathan Kallus

JMLR 2020 · Probabilistic and scientific modeling › Treatment effects · approach: Proximal and deep causal effect estimation

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
add ARD (automatic relevance determination) kernel rescaling (ablation)KOM++ with quadratic kernel · Example 3 (n=500), simple weighting estimatormarginal MSE for SATT: 0.030391 -> 0.028481Improved
add ARD kernel rescaling (ablation)KOM++ with Gaussian kernel · Example 3 (n=500), simple weighting estimatormarginal MSE for SATT: 0.029783 -> 0.029033Improved
add ARD kernel rescaling (ablation)KOM++ with Matern 3/2 kernel · Example 3 (n=500), simple weighting estimatormarginal MSE for SATT: 0.029944 -> 0.028886Improved
replace nearest neighbor matching with KOM++ (ARD, quadratic kernel)matching estimator · Example 3 (n=500), simple weighting estimatormarginal MSE for SATT: 0.047071 -> 0.028481Improved
replace 1:1 matching with KOM++ (ARD, quadratic kernel)matching estimator · Example 3 (n=500), simple weighting estimatormarginal MSE for SATT: 0.034283 -> 0.028481Improved
add cross-validated tuning (NNM++ vs NNM) (ablation)nearest neighbor matching · Example 3 (n=500), simple weighting estimatormarginal MSE for SATT: 0.047071 -> 0.042890Improved
add augmentation with OLS outcome modelKOM++ ARD with quadratic kernel · Example 3 (n=500), simple weighting estimatormarginal MSE for SATT: 0.028481 -> 0.028698No clear effect
add augmentation with OLS outcome modelNo matching · Example 3 (n=500), simple weighting estimatormarginal MSE for SATT: 0.034188 -> 0.032925Improved
replace propensity matching with KOM++ in low overlap IHDPIPW and 1:1 matching · IHDP semi-synthetic (low overlap)MSE for SATTImproved
replace Gaussian/Matern kernel with quadratic kernel (ablation)KOM++ with ARD · Example 3 (n=500), simple weighting estimatormarginal MSE for SATT: 0.029033 -> 0.028481No 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