Safeguarding Educational Institutes: An Ensemble Model for Intrusion Detection

Abstract

A beginning of digital era emerged after COVID-19 where many sectors were running towards digital enhancements to minimize the effect of pandemic. Among these sectors, the most prominent one was Education sector, which showed a significant amount of improvement from its past practices. Being digital increases the visibility of users as well as intruders ready to take advantage of any loophole. Education sector is considered as newcomer in the digital advancements as it lacks in cyber awareness among students, teachers and managerial people.It was observed that during the pandemic time, the education sector was struggling to understand the shortcomings of digital learning, digital platforms along with cyber threats. The Indian response team CERT for cybercrime, reported education sector as third largest sector prone to cyber-attacks in past few years. The age and lack of understanding of users in education sector, makes it more vulnerable to cyber-attacks. This paper, proposes the concept of ensemble learning and augment the prediction model using machine learning classifiers for intrusion detection. Our propose focuses on two classifiers decision tree and XGBoost classifier; later these two can be combined to make ensemble model.  The results of decision tree classifier has shown the accuracy of 99.86 whereas XGBoost classifier has shown accuracy of 99.91 and precision of  decision tree classifier is 99.84 whereas XGBoost classifier has 99.93. By comparing the performance of both the classifiers has shown that XGBoost classifier gives better result than decision tree.