Skip to content

Conversation

@Aman2006-code
Copy link

This PR introduces a set of improved anomaly detection functions for identifying peak edit activity in Wikimedia projects, as part of the Community Activity Alerts system.


What's Added

New Detection Methods:

  • STL + Z-score: Seasonal-Trend decomposition with residual anomaly detection.
  • Facebook Prophet: Time series forecasting + residual outlier detection.
  • Signal-based: Prominence-based peak detection using scipy.signal.

Files Added:

  • polars_migration/advanced_anomaly_detection.py: Contains all new detection functions.
  • anomaly_detection_methods.md: Research summary, comparison, and recommendation for default use (STL+Z).

🛠️ Contributed by: Aman2006-code
📅 Date: 7 August 2025

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

1 participant