Hey, for the demonstration about the Adversarial risk vs the True risk of the lecture about adversarial ML, I really don't get a lot of things:
1) If I understood well y^ is the prediction, while y is the true label, right?
2) How can we transform the product of p(Xi | y) in the product of p(Xi | y^)? I really don't get what is going on.
Adversarial - last demonstration
Hey, for the demonstration about the Adversarial risk vs the True risk of the lecture about adversarial ML, I really don't get a lot of things:
1) If I understood well y^ is the prediction, while y is the true label, right?
2) How can we transform the product of p(Xi | y) in the product of p(Xi | y^)? I really don't get what is going on.
Thanks in advance, have a good day
Hi Giorgio,
You are right. These are typos which will be corrected. I will upload a new version soon.
Thank you very much.
Cheers,
Nicolas
1
Add comment