Research map / Probabilistic and scientific modeling / Treatment effects
Modeling Interference for Treatment Effect Estimation in Network Dynamic Environment
Qiang Huang, Jin Tian
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 changed | Where | Measured | Effect |
|---|---|---|---|
| add gradient reversal layer (ablation) | DSPNET · Flickr | sqrt(PEHE): 2.179 -> 1.497 | Improved |
| add gradient reversal layer (ablation) | DSPNET · BlogCatalog | sqrt(PEHE): 1.886 -> 1.464 | Improved |
| add interference modeling module (ablation) | DSPNET · Flickr | sqrt(PEHE): 1.938 -> 1.497 | Improved |
| add interference modeling module (ablation) | DSPNET · BlogCatalog | ATE error: 1.118 -> 0.845 | Improved |
| add GRU temporal encoder (ablation) | DSPNET · Flickr | sqrt(PEHE): 10.235 -> 1.497 | Improved |
| add GRU temporal encoder (ablation) | DSPNET · BlogCatalog | sqrt(PEHE): 10.652 -> 1.464 | Improved |
| use 1 GCN layer instead of 3 (ablation) | DSPNET · Flickr | sqrt(PEHE): 15.703 -> 1.497 | Improved |
| use 1 GCN layer instead of 2 (ablation) | DSPNET · BlogCatalog | sqrt(PEHE): 6.293 -> 1.464 | Improved |
| replace GRU cell with LSTM cell (ablation) | DSPNET · Flickr (p=0.1%) | sqrt(PEHE): 1.497 -> 1.498 | No clear effect |
| add unobserved confounding at 30% level (ablation) | DSPNET · Flickr | sqrt(PEHE): 1.497 -> 1.712 | Hurt |
| increase network dynamics from p=0.1% to p=1.0% | DSPNET · Flickr | sqrt(PEHE): 1.497 -> 2.189 | Hurt |
| use 4 MLP layers instead of 3 (ablation) | DSPNET · Flickr | sqrt(PEHE): 1.455 -> 1.571 | Hurt |
Closest papers by meaning
- Higher-Order Causal Message Passing for Experimentation with Complex Interference (NeurIPS 2024 · similarity 0.91)
- Online Experimental Design With Estimation-Regret Trade-off Under Network Interference (NeurIPS 2025 · similarity 0.90)
- Independent-Set Design of Experiments for Estimating Treatment and Spillover Effects under Network Interference (ICLR 2024 · similarity 0.88)
- Doubly Robust Causal Effect Estimation under Networked Interference via Targeted Learning (ICML 2024 · similarity 0.87)
- Dynamic Treatment on Networks (NeurIPS 2026 · similarity 0.87)
- Telling Peer Direct Effects from Indirect Effects in Observational Network Data (ICML 2025 · similarity 0.87)
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.
