International Journal of Engineering and Advanced Technology (IJEAT)
Intrusion detection systems (IDS) have become a major component of network security infrastructures. Modern day intrusion detection systems are to be reliable, extensible and adaptive to the flow of network traffic and to have a low cost of maintenance. Over the years researchers have looked upon data mining as a means of enhancing the adaptability of an intrusion detection system, as it enables the IDS to discover patterns of intrusions and define valid bounds of network traffic. Despite the effectiveness of data mining based IDS it is riddled with challenges; incrementing components such as data transformations, model deployment, and cooperative distributed detection remain a labor intensive and complex engineering endeavor.