Feature Selection of Distributed Denial of Service (DDos) IoT Bot Attack Detection Using Machine Learning Techniques

Abstract

Distributed Denial of Service (DDoS) attack can be made through numerous medium and became the one of the biggest threats for computer security. One of the most effective approaches are to develop an algorithm using Machine Learning (ML). However, low accuracy of DDoS because of feature selection classifier and time-consuming detection. This research focusses on the features selection of DDoS IoT bot attack detection using ML techniques. Two datasets from NetFlow which are NF_ToN_IoT and NF_BoT_IoT are manipulated with 2 attributes selection which are Information Gain and Gain Ratio and ranked using Ranker algorithm. These datasets are then tested using four different algorithm such as Naïve Bayes (NB). K-Nearest Neighbor (KNN), Decision Table (DT) and Random Forest (RF). The results then compared using confusion matrix evaluation Accuracy, True Positive, True Negative, Precision and Recall. The result from two datasets is selected by Top 4, Top 8 and Top 12 features selection. The best overall classifier is Naïve Bayes with the accuracy of 97.506% and 90.67% for both dataset NF_ToN_IoT and NF_BoT_IoT.