Unintended Interactions Among ML Defenses and Risks - 2024

Explore the complex interplay between machine learning defenses and associated risks in this insightful 18-minute IEEE conference talk. Delve into the systematic analysis of unintended interactions that can arise when implementing multiple ML defense mechanisms simultaneously. Gain valuable knowledge about potential vulnerabilities and trade-offs in ML security strategies, and understand how these interactions may impact the overall effectiveness of defense systems. Learn from expert Vasisht Duddu as they present a state-of-the-art overview of this critical aspect of machine learning security, providing essential insights for researchers, practitioners, and security professionals in the field.