Poisoning attacks are among the greatest security threats for machine learning (ML) models. In this type of attack, an adversary tries to control a fraction of the data used to train neural networks and injects malicious data points to hinder a model's performance.
* This article was originally published here
This Blog Is Powered By Life Technology™. Visit Life Technology™ At www.lifetechnology.com Subscribe To This Blog Via Feedburner / Atom 1.0 / RSS 2.0.