AI Paper Atlas

Research map / Probabilistic and scientific modeling / Treatment effects

The Adaptive Doubly Robust Estimator and a Paradox Concerning Logging Policy

Masahiro Kato, Kenichiro McAlinn, Shota Yasui

NeurIPS 2021 · Probabilistic and scientific modeling › Treatment effects · approach: Proximal and deep causal effect estimation · 3 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 ADR estimator instead of AIPW with true logging policyADR · Section 6.1 LinUCB, T=250RMSE: 0.158 -> 0.061Improved
use ADR instead of AIPWADR · Section 6.1 LinUCB, T=750RMSE: 0.088 -> 0.033Improved
use ADR instead of AIPWADR · Section 6.1 LinTS, T=750RMSE: 0.060 -> 0.030Improved
use ADR instead of IPW with true logging policyADR · Section 6.1 LinUCB, T=250RMSE: 0.205 -> 0.061Improved
use ADR instead of EIPW (estimated logging policy)ADR · Section 6.1 LinUCB, T=750RMSE: 0.134 -> 0.033Improved
use ADR instead of DMADR · Section 6.1 LinUCB, T=250RMSE: 0.073 -> 0.061Improved
replace AIPW with AW-AIPW (StableVar weight)AIPW · Section 6.1 LinUCB, T=250RMSE: 0.158 -> 0.179No clear effect
use ADR instead of AIPWADR · mnist LinUCB, T=1000, alpha=0.4RMSE: 0.112 -> 0.028Improved
use ADR instead of AIPWADR · satimage LinUCB, T=1000, alpha=0.4RMSE: 0.019 -> 0.022Mixed
use ADR instead of AIPW at alpha=0.1ADR · mnist LinUCB, T=1000, alpha=0.1RMSE: 0.085 -> 0.086Hurt
increase sample size T from 250 to 750ADR · Section 6.1 LinUCBRMSE: 0.061 -> 0.033Improved
increase sample size T from 250 to 750EIPW · Section 6.1 LinUCBcoverage ratio (CR): 0.81 -> 0.13Hurt

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