AI based IDS & AML for IDS

We named this project IDS2, which stands for Interpretable, Dependable, and Secure Intrusion Detection System.

This website provides a comprehensive resource on AI-based Intrusion Detection Systems (IDS) and Adversarial Machine Learning for IDS. 

If you have any questions or require further information about these topics, please feel free to contact Associate Professor Dan Dongseong Kim (at dan.kim@uq.edu.au.) at The University of Queensland, Australia. 

Interpretable, Dependable, and Secure Intrusion Detection System (IDS^2)

Interpretable (or eXplainable)

• Supports both white-box and black-box models.

• Provides explanations in natural language with clear boundary conditions and visualizations.

• Enables justification, control, improvement, and discovery of security insights.

Dependable

• Ensures safety and reliability against both accidental and intentional disruptions.

Secure

• Designed to resist evasion attacks and maintain robustness against adversarial threats.

Our Research on AI Security & Security for AI