Intrusion Classifier based on Multiple Attribute Selection Algorithms

with the rapid growth of attack patterns, the number of attributes for detecting attacks gradually increased. Moreover, an automatic attack classification method, as the next thing of intrusion detection, is needed. For solving the above problems, an intrusion classifier based on multiple attribute selection algorithms has been proposed. The classifier includes various combinations with different representative attributes selection algorithms and classification algorithms. A series of experimental results on well-known KDD Cup 1999 data sets indicate real time performance and classification performances of different combinations.