There is a marked increase in transactional services such as online shopping, online trading etc. provided on the internet.[6] Along with this growth, there is an increase in the frequency of malicious attacks attempting to breach the websites’ integrity. The Denial of Service (DOS) is one type of malicious attack on the internet world which experienced infamy during the late 1990s and it’s still a cause of a problem for network security officials today. DOS attacks are rapidly developing a threat to today’s Internet. This paper explores the concept of detection of DOS attacks using various data mining algorithms such as Random Forest, KNN and SVM. It presents a comprehensive survey of DOS attacks and detection method algorithms. In this paper- open issues, research challenges and possible solutions in this area are also highlighted. NSL-KDD Cup ‘99 dataset[8] is used for applying Data Mining algorithms and testing. This paper aims at developing a system to detect DOS attacks on a system in real time and most accurately distinguish between legitimate and malicious network traffic. Keywords- DOS attack, Random Forest, KNN, SVM, NSL-KDD dataset.