An approach to the detection of distributed denial of service attacks on simulation data

Data mining is the name given to all the processes of collecting, cleaning, processing, analyzing and obtaining useful information from data. It is one of the fastest growing areas in the computer industry. It started as a small field of interest in computer science and statistics and is quickly forming its own field. Because it can be applied to large datasets, it can be used in almost all industries such as retail, manufacturing, telecommunications, healthcare, insurance and transportation. Studies on predictive modeling have been increasing recently. Modeling-based algorithms are used quite frequently, especially in the fields of health, service and informatics. The aim of this study is to create a data mining-based prediction model to detect Distributed Denial of Service attacks. This model is to predict the damage to the system when the attacks first start and to predict the success percentage of the attack. In this study, simulated data containing 500 cases were generated based on real-life distributed denial-of-service attack scenarios. The independent variables in the data set are the lifetime of the packets, the number of packets attacked, low network performance status, slowness in accessing websites, whether there are interruptions in network connections, whether there is an increase in the number of spam e-mails, access to certain parts of the website is blocked, dependent variable is the success of the attack. In this study, Random Forest method, which is frequently used in the literature and has good performance, is discussed as a classification method.