Cyber-attacks have the potential to cause power outages, issues with military hardware, and breaches of personal information, such as the theft of medical records if they fall into the wrong hands. One of the most important issues with network security in recent years has been denial of service (DoS) and distributed denial of service (DDoS). The detection and mitigation against DoS/DDoS assaults has been the topic of a lot of related tasks and techniques. There are numerous types of cyberattacks, including ransomware, phishing ... Such. are occurring more frequently. Researchers are devoting time to figuring out how to defend against these assaults. Machine Learning is greatly assisting in the training of models to identify the attack and then stop it before it causes the most harm. This study reviews numerous machine learning methods to identify DoS/DDoS attacks. If the suggested feature selection and classification algorithm are followed, it is possible to anticipate a DDoS assault in a network with a maximum accuracy of 99.83 per cent using a machine learning approach. The user is free to select either a feature selection algorithm or a classification algorithm. The sequential feature Selection (SFS) and the Information gain techniques are many of the feature selection algorithms that outperforms previous baseline approaches in terms of detection accuracy and model construction overhead. Following the decision, the web application establishes a connection to the socket and begins recording and categorizing virtual network traffic. Dynamic charts are used to convey information about attack instances (if any), attack packet count to the client after the capture is terminated. Only those attributes from the incoming packet that are required to accurately predict the class of that packet are employed by the trained model that is used to classify real-time packets.