In simulations, Bayesian GLASSO estimation outperforms frequentist GLASSO on bias in edge weights, centrality measures, correlation between estimated and true partial correlations, and specificity¶

Claim · All claims
Evidence · 1 study · causal r2 · q1
Currency · newest study 2022 · not yet checked against its sources · not reviewed by a person

Subclaims¶

q1 i? The Bayesian GLASSO outperformed the frequentist GLASSO with respect to bias in edge weights, centrality measures, correlation between estimated and true partial correlations, and specificity in extensive simulations. → Joran Jongerling 2022

Evidence¶

Joran Jongerling 2022¶

Joran Jongerling, Sacha Epskamp, Donald Williams. (2022). Bayesian uncertainty estimation for Gaussian graphical models and centrality indices. Multivariate Behavioral Research. https://www.nwea.org/research/publication/bayesian-uncertainty-estimation-for-gaussian-graphical-models-and-centrality-indices/

q1 · i? · causal · r2

Extensive simulation study comparing estimation of symptom networks with Bayesian GLASSO- and Horseshoe priors to frequentist GLASSO; results showed that "the Bayesian GLASSO outperformed the frequentist GLASSO" on the listed measures. No effect sizes are printed in the available text.

"the Bayesian GLASSO outperformed the frequentist GLASSO with respect to bias in edge weights, centrality measures, correlation between estimated and true partial correlations, and specificity."

Discussion¶

Related Claims¶

• In simulations, the Bayesian GLASSO performed better than the Horseshoe prior for estimating symptom networks — related
• Bayesian GLASSO shows good coverage of uncertainty for strength and closeness centrality, but uncertainty in betweenness centrality is estimated less well — related
• Sensitivity is better for the frequentist GLASSO than for the Bayesian GLASSO, though Bayesian GLASSO performance is usually close — related
• The frequentist graphical LASSO has difficulty estimating centrality indices and uncertainty in these measures in symptom networks — related
• Changes in ICT career aspirations are not systematically linked to contemporaneous changes in other occupational domains — related
• Regularization-based graphical modeling has key drawbacks, including difficulty obtaining valid parameter uncertainty and inflated false positive rates — a broader claim this one bears on
• Closing knowledge gaps showed the highest association with performance among all considered variables (partial r = 0.17) — related