Cross-Conformal Anomaly Detection

Supplementary code for the paper 'Uncertainty Quantification in Anomaly Detection with (Cross-)Conformal p-Values'

Uncertainty Quantification in Anomaly Detection

The provided code defines an experimental evaluation setup for comparing the classical inductive conformal (or split-conformal) approach for anomaly detection to cross-conformal methods as derived from (cross-)conformal prediction.

Setup

• Run with Python 3.12 (>=3.9)
• Install requirements.txt (pip install -r requirements.txt)

Contact

Ask questions, report bugs: oliver.hennhoefer@h-ka.de

Conformal Inference

To learn more about conformal inference, check out the awesome conformal prediction repository.