AI Paper Atlas

Research map / Probabilistic and scientific modeling / Treatment effects

Counterfactual Prediction for Bundle Treatment

Hao Zou, Peng Cui, Bo Li, Zheyan Shen, Jianxin Ma, Hongxia Yang, Yue He

NeurIPS 2020 · Probabilistic and scientific modeling › Treatment effects · approach: Benchmarking treatment effect models · 18 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
re-weight samples with variational sample weights (VSR) instead of no re-weightingDNN outcome predictor · Synthetic, p=10, n=10000test RMSE: 0.617 -> 0.476Improved
re-weight samples with VSR instead of no re-weightingDNN outcome predictor · Synthetic, n=10000, p=50test RMSE: 1.940 -> 1.506Improved
re-weight with VSR instead of no re-weightingDNN outcome predictor · RecSim, p=10 documentsclick-rate RMSE (x10^-2): 2.694 -> 1.596Improved
learn weights via variational latent distribution (VSR) instead of point-wise raw-treatment density ratio (Wraw) (ablation)re-weighted DNN · Synthetic, n=10000, p=10test RMSE: 0.528 -> 0.476Improved
aggregate density ratio over the variational posterior (VSR) instead of point-wise autoencoder latents (WAE) (ablation)re-weighted DNN · Synthetic, n=10000, p=50test RMSE: 1.520 -> 1.506Improved
use point-wise density ratio on raw treatments (Wraw) instead of no re-weighting (ablation)DNN outcome predictor · Synthetic, n=10000, p=20test RMSE: 0.997 -> 0.997No clear effect
use latent-space point-wise weights (WAE) instead of raw-treatment weights (Wraw) (ablation)re-weighted DNN · RecSim, p=20 documentsclick-rate RMSE (x10^-2): 3.258 -> 3.312Mixed
VSR weights vs other weights, effective sample size (ablation)re-weighting methods · Synthetic, n=10000, varying peffective sample sizeImproved

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