I recently published a longer piece on security vulnerabilities and potential defenses for machine learning models. Here’s a synopsis. Introduction Today it seems like there are about five major varieties of attacks against machine learning (ML) models and some general concerns and solutions of which to be aware. I’ll address them one-by-one below. Data poisoning […]