Detection of DDoS Attack using Machine Learning

In particular, Distributed Denial of Service (DDoS) attacks have become more sophisticated due to the internet’s rapid growth. It is essential for dependable internet services to identify and stop these attacks. This study combines the Naive Bayes and Decision Tree machine learning methods to distinguish between DDoS attacks and safe traffic. Through recursive partitioning, the Naive Bayes classification method provides interpretability whereas the Decision Tree classifier performs well for high-dimensional datasets. We suggest a technique that makes use of both techniques, and we evaluate how well it works using a large amount of data. After training on the same data, the algorithms identify attack patterns using the features capturing network properties that were extracted. Through the implementation of feature engineering and preprocessing methods, the interaction between DDOS and BENIGN patterns is highlighted, improving accuracy.