PeerJ Computer Science
Federated Learning for Cybersecurity in Internet of Things
In traditional machine learning, the data from different sources has to be moved to a central location, where the machine learning models will get trained to understand the patterns existing in the data. Due to the increased applications of Internet of Things (IoT)-based applications, sensitive data collected by IoT devices is being transferred to the cloud for training machine learning algorithms to understand the patterns in the data. The sensitivity of these data can attract malicious users into hacking attempts. The solution to this problem is a machine learning model which gets trained at the source of the data, instead of being trained at central locations like the cloud. Federated Learning is a recent advancement of machine learning, where, instead of moving the data to the central cloud, the machine learning model itself is moved to the source of the data. Hence, Federated Learning has the potential to solve several issues regarding cyber security in IoT based applications.
To find out more and to submit your abstract, please visit https://peerj.com/special-issues/88-fed-cyber
Dr Mamoun Alazab (Charles Darwin University) and Dr Thippa Reddy Gadekallu (Vellore Institute of Technology)