ABSTRACT
Online training has been present for over a decade, and its importance is increasing every day. Today, it has become very important due to ongoing COVID-19 pandemic. As it has been accepted widely by the educational systems;nowadays, challenges like hardware resources, network resources, software resource, and security have become more demanding. Security threats among all challenges require more researches to develop rigid systems where data of all stakeholders remain secured. ML has a proven track record to solve such problems. In terms of security, ML continuously learns by analyzing data to find patterns so unauthorized access to encrypted traffic is detected better and find insider threats to keep information safe. Here, a new system is being developed using an improved algorithm, described in proposed work. Using this new algorithm, machines are trained to identify unauthorized access attempts and stop them from stealing data even if, they are authenticated. © 2021, The Editor(s) (if applicable) and The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd.