Flooding Based DoS Attack Feature Selection Using Remove Correlated Attributes Algorithm

Flooding based DoS attack represents one of most danger attacks in computer networks. Maximizing the effectiveness of flooding based DoS Attack detection accuracy is the main concerns of many researchers. So, many of them are focusing on increasing the detection effectiveness by features reducing. However, limited research studies have concentrated on investigation the correlation between features together and its impact on DoS attack classification accuracy. Therefore and in this paper, remove correlated attributes algorithm has been proposed to select the most effective features on used in network traffic classification. Since that removing related features in a classification model minimizes the detection model error rate, It is a high likelihood that proposed model implementation increase flooding attack classification accuracy rate. In this research study, the proposed model experimental methodology and validation method has been highlighted.