AI Paper Atlas

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

Estimating Uncertainty Intervals from Collaborating Networks

Tianhui Zhou, Yitong Li, Yuan Wu, David Carlson

JMLR 2021 · Probabilistic and scientific modeling › Uncertainty and Gaussian processes · approach: Single-model predictive uncertainty · 6 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
learn f jointly (collaborating networks) instead of fixing f to uniform (ablation)CN-g · 1-d synthetic sin exampleconvergence of gof with sample size: U-g and CN-f converge much slower than CN-g, which has a small margin to T-gImproved
use theoretically optimal f (T-g) instead of learned f (ablation)g network · 1-d synthetic sin exampleinterval sharpnessImproved
use full CN instead of fixed uniform f (U-g) (ablation)g network · 1-d synthetic sin examplemedian and interval estimatesImproved
add CN-g over MSE or quantile regressionCN · 1-d synthetic overparameterized networkoverfittingImproved
use CN-g instead of DP on Weibull dataCN-g · Synthetic Weibullgof: -2.384 -> -1.843Improved
use CN-g instead of DP on Weibull dataCN-g · Synthetic Weibullcalibration error cal (%): 22.424 -> 2.158Improved
use CN-g instead of EN on Gaussian dataCN-g · Synthetic heteroskedastic Gaussiangof: -1.684 -> -1.684No clear effect
use CN-g instead of CQR on Gaussian dataCN-g · Synthetic heteroskedastic GaussianMAE: 0.728 -> 0.064Improved
add joint f learning (CN-g vs g-only) on Weibull (ablation)CN-g · Synthetic WeibullMAE: 0.053 -> 0.036Improved
add post-hoc recalibration CR to DP (ablation)DP · Airlinecalibration error cal (%): 16.16 -> 0.64Improved
add joint f learning (CN-g vs g-only) on small datasets (ablation)CN-g · CPUgof: -1.031 -> -1.053Hurt
use CN-f alone instead of CN-g (ablation)CN-f · Weibullgof: -1.843 -> -2.066Hurt

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