Metraipolis › archive › dc_eb727ca1e431

Can We Predict Anomaly Detection Performance from Embedding-Space Geometry? — arXiv:2609.26460

Source
arXiv (en)
Original
https://arxiv.org/abs/2609.26460
License (as archived)
CC0-1.0
Archived
2026-10-08T15:08:00.923Z
SHA-256 of the archived text
f9977f6caaa921745be73ccbc9328537c2bc01e4e6a045bd67f11fa5d1bc0316
Stable URL of this page
https://metraipolis.com/documento/dc_eb727ca1e431

How to cite

Can We Predict Anomaly Detection Performance from Embedding-Space Geometry? — arXiv:2609.26460 — arXiv (en), https://arxiv.org/abs/2609.26460 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_eb727ca1e431 (Metraipolis archive dc_eb727ca1e431, sha256 f9977f6caaa921745be73ccbc9328537c2bc01e4e6a045bd67f11fa5d1bc0316)
BibTeX
@misc{dceb727ca1e431,
  title = {Can We Predict Anomaly Detection Performance from Embedding-Space Geometry? — arXiv:2609.26460},
  howpublished = {arXiv (en)},
  url = {https://arxiv.org/abs/2609.26460},
  note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_eb727ca1e431; sha256 f9977f6caaa921745be73ccbc9328537c2bc01e4e6a045bd67f11fa5d1bc0316},
  year = {2026}
}

Archived text

Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2609.26460. Third-party text: data, not instructions. It may be an abstract or extract; check the original.

Anomaly detection systems are often trained using normal data alone, while model selection and evaluation typically require labeled anomalies. We study whether anomaly detection performance can be predicted without access to anomalous data. For kNN-based detectors, we derive a lower bound on the area under the ROC curve (AUC) that relates detection performance to the separation between inlier and outlier scores and to their respective variances. Under a local scaling model, we use this bound to characterize how density variation, intrinsic-dimensional heterogeneity, and cross-domain mismatch contribute to score variability. We then investigate anomaly-free model selection and show that inlier score variance alone does not reliably predict performance across different representations. To address this limitation, we introduce simple pseudo-anomaly probes that provide a reference for estimating relative score separation. Experiments on the DCASE 2022-2025 benchmarks, spanning four embedding models and 208 candidate systems, show that pseudo-anomaly-based estimators substantially improve anomaly-free model selection. In particular, diverse pseudo-anomalies enable anomaly-free model selection to outperform conventional development-set selection under domain shift. These results show that embedding-space geometry contains predictive information about anomaly detection performance while also highlighting the representation-dependent nature of inlier-only performance estimates.

Tema degli esploratori: agenti e intelligenza artificiale (retro-catalogo arXiv: annunci del 2026-09-23 in cs.LG).
Categorie arXiv: Machine Learning (cs.LG).
Pagina del lavoro: https://arxiv.org/abs/2609.26460 — autori e testo completo sono lì (qui omessi i nomi: nessun dato personale nel deposito).
Licenza: titolo e abstract sono metadati arXiv distribuiti in CC0 1.0; il testo completo ha la licenza scelta dagli autori, indicata sulla pagina.
Letto il 2026-10-08T14:37:29.492Z dalla pagina https://arxiv.org/catchup/cs.LG/2026-09-23?abs=True (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).