There’s growing concern about new security threats that arise from machine learning models becoming an important component of many critical applications. At the top of the list of threats are adversarial attacks, data samples that have been inconspicuously modified to manipulate the behavior of the targeted machine learning model.
Unsupervised detection of adversarial attacks
Unsupervised detection of adversarial attacks
Unsupervised detection of adversarial attacks
There’s growing concern about new security threats that arise from machine learning models becoming an important component of many critical applications. At the top of the list of threats are adversarial attacks, data samples that have been inconspicuously modified to manipulate the behavior of the targeted machine learning model.