Monitoring the traffic of the network for social media servers has become important task to avoid malicious traffic. DDOS attacks can crash the server by denying the service with malicious traffic. Various Machine Learning (ML) models has developed to identify the real traffic and malicious traffic based on network parameters. The aim of the research work is to study the performance of the various machine learning models on DDOS datasets. The feature selection methods are evaluated with ML models. The study concludes the high performance has achieved by using PCA feature selection technique on DDOS classification datasets