Coursera Flash Sale
40% Off Coursera Plus for 3 Months!
Grab it
Explore data poisoning attacks through interactive network visualization in this 21-minute conference talk that introduces GraphLeak, an open-source tool designed to make adversarial data manipulation visible and understandable. Learn how poisoned training data can be visualized through graph-based representations, allowing you to see the differences between clean and compromised datasets in real time. Discover how adversarial manipulation alters network structures and gain hands-on experience with an interactive tool that emphasizes accessibility and transparency in cybersecurity education. Understand the mechanics of data poisoning attacks through visual demonstrations that make complex security concepts more comprehensible, while exploring how malicious actors can compromise machine learning systems by manipulating training data.
