AI Paper Atlas

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

Reliable training and estimation of variance networks

Nicki Skafte, Martin Jørgensen, Søren Hauberg

NeurIPS 2019 · Probabilistic and scientific modeling › Uncertainty and Gaussian processes · approach: Infinite-width networks as Gaussian processes · 17 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 the full Combined variance-estimation method instead of standard NN variance trainingUCI regression models · UCI regression datasetstest log-likelihoodImproved
Use the full Combined method instead of standard VAE variance trainingVAE · MNIST, FashionMNIST, CIFAR10, SVHNtest log-likelihoodImproved
Add locality sampling (ablation)Combined variance network · Four UCI regression datasetstest log-likelihoodImproved
Add an inverse-Gamma variance distribution (ablation)Combined variance network · Four UCI regression datasetstest log-likelihoodImproved
Add mean-variance split training (ablation)Combined variance network · Four UCI regression datasetstest log-likelihoodImproved
Add the extrapolating variance architecture (ablation)Combined variance network · Four UCI regression datasetstest log-likelihoodImproved
Combine locality sampling, split training, inverse-Gamma variance, and extrapolation (ablation)Variance network · Four UCI regression datasetstest log-likelihoodImproved
Use locality sampling rather than ordinary mini-batches (ablation)Variance network · Gradient experimentsgradient sparsity and stabilityImproved
Use Combined uncertainty estimates for selecting active-learning pointsActive learning models · Boston, Superconduct, Powertest RMSE during data collectionImproved
Use Combined uncertainty estimates for selecting active-learning pointsActive learning models · Other UCI datasetstest RMSE during data collectionNo clear effect
Fit decoder variance with the Combined methodVAE · Artificial four-dimensional datagenerative sample qualityImproved

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