Improved Intrusion Detection System That Uses Machine Learning Techniques to Proactively Defend DDoS Attack

This abstract aims to provide a comprehensive analysis of the intricacies DDoS attacks, which are increasingly prevalent and malicious cyber-attacks that disrupt normal flow traffic targeted server by exponentially increasing network traffic. To secure distributed systems against intrusion detection mechanisms machine learning techniques commonly employed. The CICDDoS2019 dataset is often utilized for prevention these attacks. undergoes pre-processing split into training test datasets. Machine then predict classify attacks using dataset. protocols examined during attack SNMP, NTP, UDP, DNS. accuracy obtained comparing predicted results with algorithms such as K- Nearest Neighbor(K-NN)-96.49%, Support Vector (SVM)-79.61%, Random Forest (RF)-99.10%, Gaussian Naïve Bayes (GNB)-78.75% have been found produce high levels classification.