Machine Learning in Cybersecurity - Boon or Boondoggle

Explore the intersection of machine learning and cybersecurity in this keynote address from APPSEC CA 2017. Delve into the potential benefits and pitfalls of applying machine learning to cybersecurity challenges. Learn about the differences between supervised and unsupervised learning, the importance of quality data, and the challenges of class imbalance in security applications. Examine real-world examples, including academic research and personal anecdotes, to understand the practical implications of machine learning in cyber defense. Discover how DevOps practices can be integrated with machine learning approaches to enhance security outcomes. Gain insights from Dr. Zulfikar Ramzan, Chief Technology Officer of RSA, as he shares his expertise on leveraging data analytics and innovative technologies to protect against evolving cyber threats.

Intro
Agenda
Machine Learning
Machine Learning vs AI
Deep Blue
Supervised vs Unsupervised
Machine Learning in Cyber Security
Challenges Pitfalls
Good Data
Classifier
Class imbalance
Machine learning metrics
Academic paper example
Supervised Learning
A Personal Story
Unsupervised Machine Learning
DevOps and Machine Learning
Summary