2024-11-09 –, Hochschule München - R1.006
In our workshop, we would like to show you and explain the emerging dangers of machine learning. Together we will develop a threat model to improve the security of your machine learning applications.
We start with an introduction to the basics of machine learning and explain the typical process from data collection to model deployment. We then look at the specific security aspects of machine learning.
A central point of the workshop is the OWASP ML Security Top 10, which shows the most common security risks in the field of machine learning. You will learn how these threats can endanger your models and applications and what prevention strategies are available to minimize these risks. We will also look at practical approaches for identifying and eliminating security vulnerabilities in your ML projects at an early stage.
In the last part of the workshop, we will introduce you to the concept of threat modeling. Using examples and interactive exercises, we will work together to develop a threat model for an exemplary machine learning system. The aim is to develop a deep understanding of how you can systematically identify and defend against security threats. This workshop offers an ideal combination of theoretical knowledge and practical skills to sustainably improve the security of machine learning applications.
Agenda:
- Introduction to Machine Learning Fundamentals
- Overview of the Machine Learning Process
- Security in Machine Learning
- Exploration of OWASP ML Security Top 10
- Preventive Strategies for Machine Learning Security
- Concept and Examples of Threat Modeling
machine learning, threat modeling
Michael is a cybersecurity strategist and expert working on a wide range of product and cybersecurity topics with a background in secure software development. He is the co-founder of a security consulting firm that helps clients across industries implement product security programs, adopt DevSecOps, and achieve compliance with various standards. He believes that people and communication are at least as important and effective in moving organizations forward as tools and technology.
Initially specializing in cyber security, I have devoted myself more and more to the field of machine learning in recent years. Last year, I combined the two for the first time and conducted research in the field of penetration testing using reinforcement learning. Now I am looking for new ways to integrate machine learning in the most diverse areas of cyber security.