AI Paper Atlas

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

Fast Expectation Propagation for Heteroscedastic, Lasso-Penalized, and Quantile Regression

Jackson Zhou, John T. Ormerod, Clara Grazian

JMLR 2023 · Probabilistic and scientific modeling › Uncertainty and Gaussian processes · approach: Simulation-based inference

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 analytic integral reduction (univariate quadrature) instead of bivariate quadrature (ablation)Gaussian EP, heteroscedastic regression · Energy big datarun time (s): 630.3 -> 11.1Improved
use analytic integral reduction instead of bivariate quadrature (ablation)Gaussian EP, quantile regression · Energy big datarun time (s): 1234.5 -> 35.2Improved
use analytic integral reduction instead of bivariate quadrature (ablation)Gaussian EP, lasso regression · Energy big data, conservative settingsrun time (s): 6923.2 -> 33.3Improved
use analytic integral reduction instead of bivariate quadrature (ablation)Gaussian EP, heteroscedastic regression · Energy big data, conservative settingsrun time (s): 6994.1 -> 28.7Improved
use analytic integral reduction instead of bivariate quadrature (ablation)Gaussian EP, quantile regression · Energy big data, conservative settingsrun time (s): 6861.8 -> 78.1Improved
use analytic reduction (EP-1D) instead of bivariate quadrature (ablation)Gaussian EP · Heteroscedastic simulations and benchmarksL1 accuracy: 99.1 ± 0.0 vs 99.1 ± 0.0No clear effect
use EP-1D instead of Laplace approximationheteroscedastic regression · Energy big databeta1 L1 accuracy (%): 81.9 -> 99.1Improved
use EP-1D instead of Pathfinderheteroscedastic regression · Energy big databeta1 L1 accuracy (%): 25.5 -> 99.1Improved
increase Pathfinder returned samples from 100 to 1000 or 10,000 (ablation)Pathfinder · Energy big data, heteroscedasticbeta1 L1 accuracy (%): 25.5 -> 0.7 -> 9.8No clear effect
use EP-1D instead of MFVB when n < plasso regression · Lasso simulation Setting 3beta L1 accuracy (%): 89.9 -> 91.3Improved
use EP-1D instead of MFVBquantile regression · Quantile benchmarks IgGbeta L1 accuracy (%): 65.8 -> 99.1Improved
use EP-1D instead of MFVBlasso regression · Energy big datakappa L1 accuracy (%): 98.9 -> 77.7Hurt

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