AI Paper Atlas

Research map / Learning theory and algorithms / Sample complexity

A Data-Augmented Contrastive Learning Approach to Nonparametric Density Estimation

Chenghao Li, Yuanyuan Lin

JMLR 2026 · Learning theory and algorithms › Sample complexity · approach: Density and property estimation

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
add data augmentation (rho>0) to nonparametric noise contrastive estimation (ablation)NNCE · Model 1 density, d=10 (mean of 5 runs)Spearman rank correlation: 0.821 -> 0.875Improved
add data augmentation (rho>0) to nonparametric noise contrastive estimation (ablation)NNCE · Model 1 density, d=8 (mean of 5 runs)Spearman rank correlation: 0.884 -> 0.908Improved
add data augmentation to nonparametric noise contrastive estimation (ablation)NNCE · 2-D density modelsL2 distanceNo clear effect
increase rho from 0 to 5 (ablation)DANNCE · Model 1 density, d>=5Spearman rank correlationImproved
increase rho (ablation)DANNCE · 2-D modelsL2 distanceNo clear effect
increase rho (ablation)DANNCE · Model 1 density, d=3 and 4Spearman rank correlationNo clear effect
use DANNCE instead of RoundtripRoundtrip · Model 1 (2-D)L2 distance: .3865 -> .2129Improved
use DANNCE instead of telescoping density-ratio estimationTRE · Model 3 (2-D)L2 distance: 2.651 -> 2.463Improved
use DANNCE instead of telescoping density-ratio estimationTRE · Model 1 density, d>=8Spearman rank correlation: 0.929 -> 0.875Hurt
use DANNCE instead of telescoping density-ratio estimationTRE · Shuttle anomaly detectionprecision at k: 0.914 -> 0.970Improved
use DANNCE instead of RoundtripRoundtrip · Mammography anomaly detectionprecision at k: 0.482 -> 0.481No clear effect
use DANNCE instead of kernel-based two-stage estimatorTwo-stage · Model 1 (2-D)L2 distance: .6781 -> .2129Improved

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