OneShot: Label-Efficient Detector Weighting for Time Series Anomaly Detection
In this paper, we propose ONESHOT, which uses a single confirmed anomaly to weight the pool rather than to choose one detector from it.
ProofPaper ↗
Key points
- Time series (TS) anomaly detection (AD) lacks a universal detector, with the best choice determined by the series rather than by the algorithm.
- Existing methods therefore either select one detector for each series or average a pool of them.
- We show that a single confirmed anomaly is enough to come close to the best detector for the series.
- Across 152 univariate and 73 multivariate series, ONESHOT outperforms the strongest single detector, averaging the whole pool, and even picking the best detector on the confirmed event.
Sources (1)
- [1]OneShot: Label-Efficient Detector Weighting for Time Series Anomaly DetectionarXiv (AI, ML, NLP, CV, robotics, multi-agent) · Oct 12, 05:49 AM
In this paper, we propose ONESHOT, which uses a single confirmed anomaly to weight the pool rather than to choose one detector from it.
Time series (TS) anomaly detection (AD) lacks a universal detector, with the best choice determined by the series rather than by the algorithm.
Extractive summary: sentences quoted from the sources.
Before this
- Oct 12, 2026AReA: Adaptive Receptive-Field Attention for Efficient High-Resolution Diffusion Transformers
- Oct 12, 2026AUGUstus: Autonomous Driving VLA Model with Unified Generation and Understanding
- Oct 12, 2026Spectral-Driven Adaptive-Rank Optimizer for Large Language Model Reinforcement Learning with Verifiable Rewards
- Oct 11, 2026openai/openai-python v3.29.0
- Oct 10, 2026EviScene: Beyond "Is It AI?" for Agentic Image Factuality Assessment
- Oct 10, 2026LightVAE: Towards Compact and Efficient Video Autoencoders