Intrusion Detection in Industrial Internet of Things Network-Based on Deep Learning Model with Rule-Based Feature Selection

Table 4

Summary of performance comparison for UNSW-NB15 dataset.

Model	Performance metrics
	Accuracy (%)	FPR (%)	-score (%)	Recall (%)	Precision (%)	ROC curve (%)
Wrapper + neurotree [67]	98.38	1.62	0.984	0.980	0.989	0.998
SVM+EML+K-means [58]	95.75	1.87	0.944	0.997	0.897	0.986
GA +SVM [93]	97.3	0.017	0.966	0.997	0.938	0.981
CNN+LSTM [94]	94.12	—	0.956	0.989	0.925	0.984
Modified KNN [70]	98.7	1.3	0.992	0.996	0.988	0.998
CfsSubsetEval + GA+RuleEval+ANN [8]	98.8	1.2	0.989	0.989	0.989	0.998
Proposed model	98.9	1.1	0.989	0.998	0.967	0.989