MORA: Modeling Observed Changes for Drift-Robust Time-Series Anomaly Detection

Key points

• In non-stationary settings, distribution drift and true anomalies can cause similar local changes, making it difficult to tell whether a deviation reflects abnormality or evolving context.1
• We define this problem as temporal change disambiguation: determining whether a local deviation is explained by broader temporal evolution.1
• We introduce MORA, a drift-robust TSAD framework that reconstructs the same local target from paired short- and long-term views.1
• The reconstruction gap measures contextual support for a local deviation, and a data-dependent correction mechanism conservatively adjusts the primary local anomaly score.1

Sources (1)

• [1]MORA: Modeling Observed Changes for Drift-Robust Time-Series Anomaly Detection
  arXiv (AI, ML, NLP, CV, robotics, multi-agent) · Oct 7, 05:31 AM

  Time-series anomaly detection (TSAD) identifies deviations from patterns learned from historical data.

  In non-stationary settings, distribution drift and true anomalies can cause similar local changes, making it difficult to tell whether a deviation reflects abnormality or evolving context.

Extractive summary: sentences quoted from the sources.