AI Paper Atlas

Research map / Probabilistic and scientific modeling / Variational and generative sampling

Variational Weighting for Kernel Density Ratios

Sangwoong Yoon, Frank C. Park, Gunsu S YUN, Iljung Kim, Yung-Kyun Noh

NeurIPS 2023 · Probabilistic and scientific modeling › Variational and generative sampling · approach: Likelihood-free and graphical model estimation · 0 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
replace KDE with VWKDE in the K-L estimatorKDE-based density ratio · DAGM defect detectionmAUC: 0.734 -> 0.785Improved
replace KDE with VWKDEKDE-based density ratio · Unseen defect detectionmAUC: 0.926 -> 0.967Improved
replace KDE with VWKDEKDE-based density ratio · DAGM defect localizationmAP: 0.294 -> 0.369Improved
replace KDE with VWKDEKDE-based density ratio · Unseen defect localizationmAP: 0.849 -> 0.903Improved
use unsupervised VWKDE instead of supervised CNNVWKDE · DAGM defect detectionmAUC: 0.901 -> 0.785Hurt
use unsupervised VWKDE instead of supervised CNNVWKDE · Unseen defect detectionmAUC: 0.809 -> 0.967Improved
use VWKDE instead of KDE for 1D Gaussian log density ratioKDE · 1D Gaussians, 1000 samplesbias of LPDRImproved
use VWKDE instead of KDE for 1D Gaussian LPDRKDE · 1D Gaussiansvariance of LPDRHurt
vary bandwidth h for K-L estimate (ablation)VWKDE · 1D GaussiansK-L divergence estimateNo clear effect
use model-based VWKDE with Gaussian model on 20D NH GaussiansVWKDE · 20D non-isotropic heteroscedastic GaussiansK-L divergence accuracyImproved
use misspecified Gaussian score model on Gaussian mixtures (ablation)VWKDE · 2D and 20D Gaussian mixturesK-L divergence accuracyImproved

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