ML NETWORK INTRUSION DETECTION

Machine Learning / Cybersecurity

Intrusion detection on NSL-KDD and TON-IoT


WHAT IS IT?

An end-to-end machine learning pipeline for network intrusion detection using NSL-KDD and TON-IoT datasets.

NOTE

This entry documents applied machine learning research work.

DETAILS

  • > Benchmarked multiple machine learning classifiers
  • > Used ensemble methods
  • > Applied SMOTE to address severe class imbalance
  • > Performed feature selection
  • > Used hyperparameter optimisation
  • > Evaluated classification performance
  • > Improved rare privilege-escalation attack recall from 4% to 10.5%

TECHNOLOGIES

Python · Scikit-learn · Pandas · NumPy · XGBoost · SMOTE

RESULTS

  • > 78.42% accuracy with an optimised ensemble
< BACK TO PROJECTS