An Intrusion Detection System using GA and SVM classifier for IoTs

With the emergence of technologies and the increasing number of users of the Internet, such as the Internet of Things (IoT) paradigm, there are new and modern efforts to invade networks and computer systems. Many researchers have developed various artificial intelligence-based algorithms to detect these attacks. For network security they focus on machine learning models that are used in IoT and intrusion detection. In this study, we have proposed a machine learning based intrusion detection system which is a combination of support vector machine (SVM) and Genetic Algorithm (GA). GA is basically used for feature selection and parameter optimization of SVM models. The performance of a GA-based SVM model is evaluated on an KDD Cup 99 intrusion database. First, GA optimizes the KDD Cup 99 dataset and then optimizes the weight and parameters of the SVM model. The experimental result presents that the SVM model gives prominence in terms of detection rate, accuracy, false positive rate and false negative rate and also compared to other literature works.