[Dailydave] Machine Learning and Dimensions and stuff
HalVar at gmx.de
Sat Nov 22 05:34:56 EST 2014
>The implications are though, that even if the adversary adapts, that the ML analytic is forcing the adversary to operate in a smaller space to avoid appearing >anomalous. I consider anything that can shift the balance of cost from the defender to the adversary to be wildly successful.
No objection that ML-based detections will shift the burden of changing something to the attacker. I just predict that long-term success is heavily dependent on being able to keep access to the actual classifier out of the attackers' hands.
More information about the Dailydave