Research map / Probabilistic and scientific modeling / Treatment effects
A Two-Stage Penalized Least Squares Method for Constructing Large Systems of Structural Equations
Chen Chen, Min Ren, Min Zhang, Dabao Zhang
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 |
|---|---|---|---|
| use ridge regression instead of adaptive-lasso instrument selection at stage one (2SPLS vs 2SAL) (ablation) | two-stage network estimator · Simulated networks with 3 EEs, dense | FDR: over 0.8 (2SAL) | Improved |
| use 2SPLS instead of 2SAL for runtime (ablation) | two-stage network estimator · Simulated networks, n=100 | running time: 1303 vs 3239 seconds (acyclic sparse, 1 EE) | Improved |
| use 2SPLS instead of AL | adaptive lasso baseline · Simulated acyclic networks | power: under 10% -> over 80% at small sample size | Improved |
| use 2SPLS instead of SML | sparsity-aware ML · Simulated networks, n=100 | running time: 258875 -> 1303 seconds (acyclic sparse, 1 EE) | Improved |
| use 2SPLS instead of SML for dense acyclic, small n | sparsity-aware ML · Simulated dense acyclic networks, n<200 | FDR: over 40% (SML) | Improved |
| use 2SPLS instead of SML for dense cyclic networks | sparsity-aware ML · Simulated dense cyclic networks | power | Improved |
| increase sample size to large | 2SPLS vs SML and 2SAL · Simulated acyclic networks | power | No clear effect |
| use AL instead of 2SPLS for sparse acyclic network with 1 EE, moderate/large n | 2SPLS · Simulated sparse acyclic, 1 EE | FDR | Hurt |
| double error variance (ablation) | 2SPLS · Simulated networks | FDR: slightly higher FDR | Hurt |
| quadruple error variance (ablation) | 2SPLS · Simulated networks | FDR: higher FDR | Hurt |
| use t(3) errors instead of normal (ablation) | 2SPLS · Simulated networks | power: similar power, slightly increased FDR | No clear effect |
| add hub nodes to the network (ablation) | 2SPLS · Simulated sparse networks | FDR: decreased FDR in sparse networks | Mixed |
Closest papers by meaning
- Nonparametric Instrumental Variable Regression through Stochastic Approximate Gradients (NeurIPS 2024 · similarity 0.86)
- PCM Selector: Penalized Covariate-Mediator Selection Operator for Evaluating Linear Causal Effects (AAAI 2025 · similarity 0.86)
- Score-Based Causal Discovery of Latent Variable Causal Models (ICML 2024 · similarity 0.85)
- Data-Driven Selection of Instrumental Variables for Additive Nonlinear, Constant Effects Models (ICML 2025 · similarity 0.85)
- CoLiDE: Concomitant Linear DAG Estimation (ICLR 2024 · similarity 0.84)
- Latent Variable Causal Discovery under Selection Bias (ICML 2025 · similarity 0.84)
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.
