Building an Effective Intrusion Detection System by Using Hybrid Data Optimization Based on Machine Learning Algorithms

Table 6

Confusion matrix of all classes over the UNSW-NB15 dataset using DO_IDS.

Actual	Predicted										Recall
	1	2	3	4	5	6	7	8	9	10
1	35778	9	298	131	17	39	510	0	217	1	0.967
2	84	18291	419	2	1	41	0	9	21	3	0.969
3	716	13	7381	5	280	1832	166	580	154	5	0.663
4	2613	5	175	2307	6	857	12	60	27	0	0.381
5	83	1	223	0	2867	171	20	85	46	0	0.820
6	254	6	1128	3	55	1887	133	554	68	1	0.461
7	178	0	41	0	0	380	41	37	0	0	0.061
8	127	0	48	0	0	166	0	235	7	0	0.403
9	65	0	14	0	1	3	0	0	295	0	0.780
10	4	1	3	0	0	0	0	0	1	35	0.795
Precision	0.897	0.998	0.759	0.942	0.888	0.351	0.046	0.151	0.352	0.778