From the course: Complete Guide to Cybersecurity: A Practical Approach

Unlock this course with a free trial

Join today to access over 24,400 courses taught by industry experts.

Exploring data poisoning attacks

Exploring data poisoning attacks

From the course: Complete Guide to Cybersecurity: A Practical Approach

Start my 1-month free trial Buy for my team

Exploring data poisoning attacks

“

- Let's go over the different data poisoning attacks against AI systems. You have targeted attacks. And these attacks basically, is whenever the attacker aims to change their model behavior for specific inputs. For example, the attacker may want to ensure that their own face is never recognized by a facial recognition system. So they poison the training data to be able to misclassify their face. Another type of data poisoning attack is the exploratory attack. And these attacks are whenever the attacker tries to degrade the overall performance of the machine learning model or the AI model rather than actually targeting the specific inputs. So for example, you may want to actually add noise to the many types of images in, you know, visual recognition model for example. And then cause that model to plummet. Basically the accuracy of that model to plummet and to be providing false results. This is from a high level, a very high level, the steps of a data poisoning attack. Let me make that…

Contents