AI Paper Atlas

Research map / Probabilistic and scientific modeling / Uncertainty and Gaussian processes

Empirical Priors for Prediction in Sparse High-dimensional Linear Regression

Ryan Martin, Yiqi Tang

JMLR 2020 · Probabilistic and scientific modeling › Uncertainty and Gaussian processes · approach: Regression uncertainty quantification · 16 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 EB empirical prior instead of horseshoesparse linear regression prediction · Very-sparse simulationsMSPE: 0.86 vs 0.89No clear effect
use EB instead of horseshoe at moderate densitysparse linear regression prediction · Less-sparse simulations (p=125, Delta=0.2)MSPE: 2.53 -> 1.43Improved
use EB instead of lassosparse linear regression prediction · Very-sparse simulationsMSPE: 1.18 -> 0.86Improved
use EB instead of adaptive lasso at large signalsparse linear regression prediction · Very-sparse simulations (p=125, A=8, r=0.8)MSPE: 8.99 -> 1.22Improved
use EB instead of HS in the hardest settingsparse linear regression prediction · Less-sparse p=500, Delta=0.05, A=8, r=0.2MSPE: 6.44 -> 1.57Improved
use EB instead of HS for interval length95% prediction intervals · Less-sparse (p=125, Delta=0.2)interval mean length: 5.59 -> 4.20Improved
use EB instead of HS for interval coverage95% prediction intervals · Less-sparse p=500, Delta=0.05, A=8, r=0.5coverage: 0.93 -> 0.88Hurt
use EB instead of HS for very-sparse interval length at p=50095% prediction intervals · Very-sparse p=500interval mean length: 4.12 vs 4.10Mixed
use EB instead of HS for computationposterior sampling · Simulationsrun-time: about 20% of HS run-timeImproved
use discrete selection prior EB instead of continuous shrinkage HS when signals are tinyNCI-60 gene B1 · NCI-60 multidrug, B1MSPE: 0.54 -> 0.73Hurt
use EB instead of lasso/HS on real gene responsesNCI-60 genes A2, A8, A12 · NCI-60 multidrugMSPE: 0.93 vs 0.94No 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