What does this research mean for the field?

A distributed multiple testing procedure using debiased estimators can asymptotically control the directional false discovery rate for conditional dependence in Gaussian graphical models without requiring centralized data aggregation. Novelty: ClaimNovelty.METHODOLOGICAL. Consensus alignment: ConsensusAlignment.NEUTRAL.

What question did this study set out to answer?

Develop a multiple testing framework to infer conditional dependence in Gaussian graphical models across distributed datasets while controlling directional error rates and protecting privacy.

August 26, 2026Journal of Computational and Graphical Statistics

Directional false discovery rate control via distributed procedures for Gaussian graphical models

View Full Paper

Ask AI

Bookmark

Authors

ENEnsiyeh Nezakati EPEugen Pircalabelu

Discussion

Member takes

Overview

Statistical modeling study demonstrates directional false discovery rate control in distributed data settings, suggesting reliable network inference without centralizing private datasets.

Key Points

Develop a multiple testing framework to infer conditional dependence in Gaussian graphical models across distributed datasets while controlling directional error rates and protecting privacy.
Constructed multiple test statistics using debiased and distributed estimators derived from K separate local datasets without central data aggregation.
Evaluated asymptotic directional false discovery rate control and statistical power via theoretical analysis, numerical simulations, and real-world application.
Proves asymptotic control of the directional false discovery rate at any pre-specified target level under regularity conditions.
Achieves an asymptotic power of one for identifying non-zero precision matrix entries and their signs, confirmed across simulation scenarios and real data.

Cite This Study

Nezakati et al. (2026) studied this question.

synapsesocial.com/papers/6a8ebb54451774b83f3b4b6e https://doi.org/10.1080/10618600.2026.2719786

View Full Paper

Ask AI

Bookmark

August 26, 2026Journal of Computational and Graphical Statistics

Directional false discovery rate control via distributed procedures for Gaussian graphical models

View Full Paper

Ask AI

Bookmark

Authors

ENEnsiyeh Nezakati EPEugen Pircalabelu

Discussion

Member takes

Overview

Statistical modeling study demonstrates directional false discovery rate control in distributed data settings, suggesting reliable network inference without centralizing private datasets.

Key Points

Develop a multiple testing framework to infer conditional dependence in Gaussian graphical models across distributed datasets while controlling directional error rates and protecting privacy.
Constructed multiple test statistics using debiased and distributed estimators derived from K separate local datasets without central data aggregation.
Evaluated asymptotic directional false discovery rate control and statistical power via theoretical analysis, numerical simulations, and real-world application.
Proves asymptotic control of the directional false discovery rate at any pre-specified target level under regularity conditions.
Achieves an asymptotic power of one for identifying non-zero precision matrix entries and their signs, confirmed across simulation scenarios and real data.

Cite This Study

Nezakati et al. (2026) studied this question.

synapsesocial.com/papers/6a8ebb54451774b83f3b4b6e https://doi.org/10.1080/10618600.2026.2719786

View Full Paper

Ask AI

Bookmark