AI Paper Atlas

Research map / Label-efficient and robust learning / OOD and anomaly detection

Neighborhood Structure Assisted Non-negative Matrix Factorization and Its Application in Unsupervised Point-wise Anomaly Detection

Imtiaz Ahmed, Xia Ben Hu, Mithun P. Acharya, Yu Ding

JMLR 2021 · Label-efficient and robust learning › OOD and anomaly detection · approach: Anomaly detection with contaminated data · 14 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 offline NS-NMF instead of NMF/GNMF/SNMFNS-NMF · 20 anomaly benchmarksmean relative rank: 1.2 vs 3.1, 2.6, 3.1Improved
add neighborhood structure to NMF (NS-NMF vs NMF) (ablation)NMF · 20 anomaly benchmarksFriedman post hoc p-value: 7.11\times 10^{-21}Improved
add neighborhood structure (NS-NMF vs GNMF) (ablation)GNMF · 20 anomaly benchmarksFriedman post hoc p-value: 3.40\times 10^{-15}Improved
use offline instead of online NS-NMFNS-NMF · 20 anomaly benchmarkstrue positive detections: superior in 15 out of the 20 casesImproved
NS-NMF versus LoMST non-NMF methodNS-NMF · 20 anomaly benchmarkstrue positive detections: 14 cases out of 20 won, LoMST better in five cases, one tieMixed
NS-NMF vs LoMST on WaveformNS-NMF · Waveformtrue positive detections: 1 -> 9Improved
NS-NMF vs LoMST on KDDNS-NMF · KDDtrue positive detections: 39 -> 102Improved
NS-NMF vs LoMST on PageblocksNS-NMF · Pageblockstrue positive detections: 32 -> 21Hurt

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