AI Paper Atlas

Research map / Probabilistic and scientific modeling / Causal discovery

Federated Learning of Generalized Linear Causal Networks

Qiaoling Ye, Arash A. Amini, Qing Zhou

TPAMI 2024 · Probabilistic and scientific modeling › Causal discovery · approach: Differentiable and score-based DAG learning · 28 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
increase sample size from n=100p to n=10,000DARLS · Logistic GLDAG AsiaSHD: 4.6 -> 3.5Improved
add distributed data over K=20 machines instead of pooled dataDARLS · Logistic GLDAGSHDNo clear effect
split the score computation over K=2,5,10 machines instead of K=1DARLS · f(pi) on seven networksrelative loss: order of 10^{-13}No clear effect
increase number of machines K beyond 10DARLS · Insurance computation timecomputation timeNo clear effect
add the refinement step via thresholding beta-hat (ablation)DARLS · Logistic GLDAGSHDImproved
use threshold-Gaussian data instead of logistic data (model misspecified)DARLS vs HC · Threshold-Gaussian AlarmSHD: 39.8 vs 36.9Hurt
use multinomial data instead of logistic data (model misspecified)DARLS vs HC · Multinomial Hepar2SHD: 108.1 vs 96.8Hurt
lower NOTEARS penalty to 10^-4 from default 10^-1NOTEARS · Logistic GLDAG AsiaSHD: 11.8 and 10.2No clear effect
use DARLS on ChIP-Seq distributed data vs combined dataDARLS · ChIP-Seq TF networkNLR: 0.822 -> 0.701Mixed

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