… the opportunities and challenges of defending against ML-equipped inference … inference
attacks in online social networks as an example to illustrate the opportunities and challenges. …

… the works for the membership inference attacks against federated machine learning-based …
the state-of-the-practice of system security for the machine learning-based software …

… learning-based or … protection the noise introduced to the model decreases model accuracy
to 35 percent. This is a significant challenge in the mitigation of membership inference attacks…

… various issues and challenges of ML security mechanisms … They demonstrated the procedure
of machine learning based network … inference attack is discussed at the end of this section. …

… open issues that affect security systems based on machine … for more robust machinelearning-based
techniques that can … , he does not try to infer the existence of a similar function by …

… We discuss the challenges of privacy attacks on cloud-based … Attack) security, it is not
necessary to discuss privacy issues … Whereas, directly performing a fully-learning-based attack …

… the security issues of machine learning, focusing on existing … and membership inference
attack can steal the model param… to compromise these deep learning based security applications…

… the data security attacks to deep-learning-based systems (… is to destroy the availability of the
output during the training process, … the inference service to collect the user’s sensitive data. …

… applications, ML also raises unique challenges in terms of security [2ś4]. In particular, … the
training data in and relects it in the model’s output behavior. Thus, we can infer the training …

… of security of IoT in terms of different types of possible attacks. Moreover, ML-based
potential solutions for IoT security has been presented and future challenges are discussed. …