Design network intrusion detection system using support vector machine

The growing use of the Internet and the existence of vulnerable points in networks have made the usage of intrusion detection systems as one of the most important security elements. This study aimed to present a method to design an analytical framework of detecting destructive data with respect to three factors including time, users' information, and scale. The design can be applied for big data. In the proposed method, to train data, the time has been divided into subperiods exploiting users' review information during each period of time, and the data have been trained. Also, storing methods have been applied for scalability to enhance the speed and reduce the volume of computations. The method used in this study is a combination of hardware-software method to detect destructive data to cluster them (VIRUS TOTAL Dataset). Also, the proposed method applied a new algorithm of modified vector machine, and the efficiency of the algorithm has promoted support vector machine (SVM), designed to operate better than previous methods. The results showed that the proposed method is more acceptable than other previous methods. The results indicated that the method works with the accuracy of 0.97 which can be fairly accepted.