AI Paper Atlas

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

Unsupervised Anomaly Detection with Rejection

Lorenzo Perini, Jesse Davis

NeurIPS 2023 · Label-efficient and robust learning › OOD and anomaly detection · approach: Anomaly detection with contaminated data · 2 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
Set rejection using ExCeeD stability confidence and a constant threshold instead of the tested baselinesUnsupervised anomaly detectors · 34 benchmark datasets, averaged over detectorscost per example: 0.121 ± 0.133 vs 0.125 ± 0.135 (SS-Repen); 0.148 ± 0.144 (NoReject)Improved
Use ExCeeD-based rejection rather than metric-based threshold selectionUnsupervised anomaly detectors · 34 benchmark datasetscost per example: more than 5% vs SS-Repen; 11% vs Ens; 13% vs Mv and Udr; 17% vs Em; 19% vs NoRejectImproved
Use RejEx rather than the baselines when false positives are costlyUnsupervised anomaly detectors · 34 benchmark datasets; false positive cost = 10cost per example: 11 out of 12 detectorsImproved
Use RejEx rather than the baselines when false negatives are costlyUnsupervised anomaly detectors · 34 benchmark datasets; false negative cost = 10cost per example: 11 detectorsImproved
Use RejEx rather than the baselines when rejection cost is lowUnsupervised anomaly detectors · 34 benchmark datasets; false positive cost = 5, false negative cost = 5, rejection cost = γcost per example: all detectorsImproved
Replace label-free thresholding with label-optimized thresholdingExCeeD confidence · 34 benchmark datasets, averaged over detectorscost per example: 0.121 ± 0.133 vs 0.121 ± 0.133; 0.6% higher costNo clear effect
Use RejEx instead of NoRejectIForest · 34 datasetsCPU time per training example (ms): 0.0±0.0 -> 0.06±0.22Hurt
Use RejEx instead of threshold-search baselinesIForest · 34 datasetsCPU time per training example (ms): 90±68 (SS-Repen); 916±900 (Stability); RejEx: 0.06±0.22Improved

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