Abstract:

Denial of service attack is a common cyber attack method that is difficult to detect and prevent for a long term. By consuming the bandwidth or computing resources of the target computer, the target computer network service is interrupted or stopped, which results in the normal users can not access it. With the rapid development of machine learning algorithms, decision tree, support vector machine, random forest and adaboost are gradually used to identify and detect DoS attacks network traffic. For most machine learning algorithms, the choice of network traffic characteristics directly determines the performance of the algorithm. This paper extracts and selects network traffic characteristics by using CICFlowMeter and random forest algorithm, and designs algorithm training model to detect DoS attack traffic, which achieves better accuracy and recall rate, and verifies the validity of the detection method.

Key words: DoS attack, machine learning, random forest, feature selection, ensemble learning