Skip to main content
. 2016 Oct 13;16(10):1701. doi: 10.3390/s16101701

Table 1.

The distribution of the training and testing sets from the six datasets generated from KDD’99 and NSL-KDD.

Dataset Training Dataset Testing Dataset
Normal% DoS% Probe% U2R% R2L% Normal% DoS% Probe% U2R% R2L%
Dataset 1 17.96 72.28 7.583 0.096 2.079 19.48 73.90 1.339 0.073 5.205
Dataset 2 19.48 78.40 1.645 0.021 0.451 19.48 73.90 1.339 0.073 5.205
Dataset 3 19.69 79.23 0.831 0.011 0.228 19.48 73.90 1.339 0.073 5.205
Dataset 4 53.38 36.65 9.086 0.044 0.860 43.07 33.08 10.73 0.887 12.21
Dataset 5 48.56 33.11 16.81 0.075 1.435 43.07 33.08 10.73 0.887 12.21
Dataset 6 53.38 36.65 9.086 0.044 0.830 18.16 36.64 20.27 1.688 23.24