|Project Description: ||Due to increase in computer usage and computer based information systems, the security of networks is considered as a primary issue. The computers are being exposed to several new attacks and malicious intrusions over the internet. Intrusion detection systems (IDS) are in demand as these are one of the effective ways for detecting malicious attacks and intrusions. Intrusion detection systems are being widely used to preserve the confidentiality, integrity and availability of the information. Various Data mining techniques are being used for intrusion prevention and detection by analyzing large network traffic. Data mining techniques are used for processing large volumes of data and to detect unknown patterns of the attackers. There are various algorithms for intrusion detection based on data mining. Most of the existing methods suffer from low accuracy and high false alarm rate. The proposed solution overcomes these drawbacks by improving the accuracy and reducing the false alarm rate. The system combines the clustering and classification techniques thereby forming a hybrid approach.