Explainable and Uncertainty-Aware Machine Learning for Reliable Network Intrusion Detection: Calibration, Conformal Prediction, and Cross-Dataset Generalization
Code and measurement logs for the paper "Explainable and Uncertainty-Aware Machine Learning for Reliable Network Intrusion Detection: Calibration, Conformal Prediction, and Cross-Dataset Generalization." The scripts reproduce the main results (RQ1-RQ6): detection, imbalance-aware calibration (attack-region ECE), split/Mondrian conformal prediction, SHAP explainability, and a confound-free cross-dataset generalization study on the unified NetFlow v2 datasets (NF-UNSW-NB15-v2, NF-CSE-CIC-IDS2018-v2). Raw datasets are not included; see README for sources. Licensed under MIT. Version 1.0.1 adds scripts missing from 1.0.0 (see README, "Changes in
Resource information
- Format
- zip
- Original repository
- 23399891
- DOI
- 10.5281/zenodo.23399891
- Credit basis
- Creators in Zenodo record metadata
- Access
- Open files available from the original Zenodo deposit
- Last checked
- Nov 5, 2026
- License
- mit-license
- Last record update
- Nov 5, 2026
About this resource
Code and measurement logs for the paper "Explainable and Uncertainty-Aware Machine Learning for Reliable Network Intrusion Detection: Calibration, Conformal Prediction, and Cross-Dataset Generalization." The scripts reproduce the main results (RQ1-RQ6): detection, imbalance-aware calibration (attack-region ECE), split/Mondrian conformal prediction, SHAP explainability, and a confound-free cross-dataset generalization study on the unified NetFlow v2 datasets (NF-UNSW-NB15-v2, NF-CSE-CIC-IDS2018-v2). Raw datasets are not included; see README for sources. Licensed under MIT. Version 1.0.1 adds scripts missing from 1.0.0 (see README, "Changes in version 1.0.1").