Machine learning is a powerful new tool that can be used for security applications (for example, to detect malware) but machine learning itself introduces many new attack surfaces. For example, attackers can control the output of machine learning models by manipulating their inputs or training data. In this session, I give an overview of the emerging field of machine learning security and privacy.
Learning Objectives:
1: Learn about vulnerabilities of machine learning.
2: Explore existing defense techniques (differential privacy).
3: Understand opportunities to join research effort to make new defenses.
Speaker: Ian Goodfellow
Ian Goodfellow is a Staff Research Scientist. He leads a research group in Google Brain studying adversarial techniques in AI. He is the Lead Author of the first major textbook on deep learning—Deep Learning (MIT Press). With his collaborators at Google, he published some of the first research on security and privacy of deep learning.
Detailed Presentation:
Comments