Performance Analysis of Indicative Feature Subset Selection Algorithm for Effective Intrusion Detection

Abstract

            Intrusion detection and prevention is considered as a foremost process in today's digital world to provide security to handheld devices and other systems. There are many ways in which we can identify the intrusions but the existing approaches are lacking performance in identifying mobile intrusions due to its varying nature. Some application parameters like location and system properties which are not fixed are causing serious threats to the user's personal data. Hence we need an efficient approach which can improve the accuracy in addressing the above challenge. We have developed an framework namely Indicative Feature Subset Selection (IFSS)-SVM to meet the above challenge. By considering the data collected from mobile devices we have checked the performance of existing algorithms with our approach and we have found an good improvement which is discussed in the paper.