Hybrid Intrusion Detection System using K-means and K-Nearest Neighbors Algorithms

With the widespread use of technology, companies of all sizes have benefited from the use of resources and technologies on the Internet. On the other hand, real security threats are increasing problems and intrusion detection systems (IDS) can help protect external and internal organizations and provide network security. The intrusion detection system only monitors network traffic and informs the system administrator for unusual activity. It is very similar to a home alarm system that will turn on the alarm when the thief enters the window or door. Various methods such as machine learning, data processing and optimization are also analyzed to support the important developments of IDS and to assist better future problems suggestions. In this paper we use hybrid data mining methods such as k-means and k-nearest neighbors to reduce time complexity of the system with great accuracy. This model is implemented by using KDD’99 data set.