Research into the network security model blended of data stream mining and intrusion detection system

In response of the fact that traditional intrusion detection systems are not able to fulfill the requirements for specific network security, such as fast processing speed, stronger defense capability, and higher real-time performance, a model of network security defense is built on the integration of data stream mining and intrusion detection; and, a data stream clustering algorithm is designed for mining in the model. Through analysis and simulation, the model turns out to be higher in detection rate and lower in false-alarming or false negative rate, thus achieving a better result.