A Detailed Analysis of the KDD CUP 99 Data Set

Abstract

During in last decade, anomaly detection It has attracted the attention of many researchers to overcome the weakness of signature-based IDS in detecting new attacks, and KDDCUP'99 is the most widely used dataset for grade from these systems. Having held a statistical analysis of this data set, we found two important issues that greatly affects the performance of the systems being evaluated, and the resultsin a very poor evaluation of anomaly detection approaches. To to solve these problems, we proposed a new dataset, NSL-KDD, which the consists from selected records from in full KDD datainstalled as well as does No suffer from Any from mentioned limitations