CS X410.D - Artificial Intelligence for Cybersecurity

Course Description

Artificial Intelligence (AI) and machine learning (ML) technologies and techniques are being deployed for use in cybersecurity. These technologies include network anomaly detection, biometric authentication, data analytics to uncover fraud as well as spam detection. Attackers leverage advanced ML algorithms to gain an advantage on their potential targets. ML systems are susceptible to adversarial input perturbations impacting deep neural networks. It is critical to develop and deploy secure ML systems integrating software security best practices. This course will facilitate an understanding of Adversarial Machine Learning (AML), key types of attacks, defenses as well as fundamental properties for explainable AI systems. This course will examine the unique challenges for user trust in AI systems, identifying and managing bias in AI systems as well as critical issues related to AI and ML impact on cybersecurity.

Learner Outcomes

• Articulate a taxonomy and terminology of Adversarial Machine Learning (AML).
• Articulate a conceptual hierarchy including key types of attacks, defenses, and consequences.
• Articulate the four principles for explainable artificial intelligence that comprise the fundamental properties for explainable AI systems.
• Identity the five categories of explainable AI and major classes of explainable algorithms.
• Identify the distinct requirements for machine learning systems as well as experimental psychology pertaining to interpretation and comprehension.
• Articulate the unique challenges for user trust in AI systems.
• Identify and manage bias in AI systems.
• Articulate the critical issues related to AI and ML, specifically AI \ ML impact on cybersecurity.

Applies Towards the Following Certificates

• Applied Cybersecurity : Elective