Security Threats to Machine Learning Systems
Rohit Soni, Sparsh Paliya, Lalita Gupta · 2022
Machine Learning not only promises to solve issues, but it also has the potential to assist businesses in formulating forecasts and so enhancing decision-making. Machine learning, on the other hand, has security difficulties. This study examines the many security concerns that machine learning faces and how they might be mitigated. There are two types of security risks to ML systems: those that occur before model training and those that occur after model training. The document also offers information on a variety of additional threats that fall under these categories and target other sorts of data while posing varied hazards.