Reproduce “MORA: Modeling Observed Changes for Drift-Robust Time-Series Anomaly Detection” on OpenResearch
Turn your favorite coding agents into research agents that review literature, develop hypotheses, and run experiments.
Universal app for Apple silicon and Intel. Requires macOS 11 or later.
About this paper
Xudong Mou, Tiejun Wang, Rui Wang, Hui Wang, Pin Liu, Tianyu Wo, Xudong Liu, Renyu Yang
MORA detects time-series anomalies by correcting local scores with temporal context. It reconstructs the same short-window target from local and longer views, subtracting a learned fraction of the positive reconstruction gap only when the contextual reconstruction is better. On SMD, its Best RPA-F1 is 69.97%, versus 60.74% for AOC; this result does not establish that it identifies the cause of changes.
Summary by alphaXiv.
Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.