AI Paper Atlas

Research map / Probabilistic and scientific modeling / Causal discovery

The Reduced PC-Algorithm: Improved Causal Structure Learning in Large Random Networks

Arjun Sondhi, Ali Shojaie

JMLR 2019 · Probabilistic and scientific modeling › Causal discovery · approach: Differentiable and score-based DAG learning · 17 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 rPC-approx instead of PC-AlgorithmPC-Algorithm · Power law DAGstrue vs false positive rate (pROC)Improved
use rPC instead of PC-AlgorithmPC-Algorithm · Erdos-Renyi DAGstrue vs false positive rate (pROC)No clear effect
use rPC-approx instead of rPC-full (exhaustive separating-set search)rPC · simulated DAGsaccuracyNo clear effect
use rPC-approx instead of PC-AlgorithmPC-Algorithm · Power law p=500 n=200runtime ratio: 100 -> 0.36Improved
use rPC-full instead of PC-AlgorithmPC-Algorithm · Erdos-Renyi p=500 n=200runtime ratio: 1 -> 23.25Hurt
use rPC-approx instead of PC-AlgorithmPC-Algorithm · Erdos-Renyi p=100 n=200runtime ratio: 0.98No clear effect
tune eta by BIC over 1 to 4 in rPC-approxrPC-approx · power law DAGstrue vs false positive rate (pROC)Improved
use rPC-approx instead of PC-AlgorithmPC-Algorithm · Prostate cancer TCGA networkhub genes identified: 6 -> 19Improved
path faithfulness instead of restricted strong faithfulnessPC-Algorithm assumption · Power law, expected degree 2, p=20probability condition satisfied: 0.54 -> 0.85Improved
path faithfulness instead of restricted strong faithfulnessPC-Algorithm assumption · Erdos-Renyi, expected degree 5, p=10probability condition satisfied: 0.10 -> 0.61Improved
path faithfulness instead of restricted strong faithfulnessPC-Algorithm assumption · dense graphs, p=30probability condition satisfied: 0 -> 0.01No 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