Published November 14, 2026
| Version v1.0.0
Software
Open
When 100% Accuracy Is a Warning Sign: analysis code for a condition-blocked audit of ML leak detection
Authors/Creators
- 1. Memorial University of Newfoundland
- 2. Hamad Bin Khalifa University
- 3. University M'Hamed Bougara of Boumerdes
Description
Analysis and figure-generation code for "Evaluating Near-Perfect Accuracy Claims in Machine-Learning-Based Leak Detection for Two-Phase Pipelines: A Condition-Blocked Audit" (Suleiman, Serpoosh, Elkholy, Ferroudji, Rahman, Hamilton).
The paper is the reference for all reported numbers; this release contains only code. Dependencies are pinned in requirements.txt (Python 3.11.3).
Notes
Files
issahnazif-afk/100-percent-warning-sign-v1.0.0.zip
Additional details
Related works
- Is supplement to
- Software: https://github.com/issahnazif-afk/100-percent-warning-sign/tree/v1.0.0 (URL)
Software
- Repository URL
- https://github.com/issahnazif-afk/100-percent-warning-sign