Employing a Machine Learning Approach to Detect Combined Internet of Things Attacks against Two Objective Functions Using a Novel Dataset

Table 9

Experiment 4 results.

	RSME	MAPE (%)	ROC average	Correctly classified instances (%)	Balancing technique average (%)
SMOTE–NB	0.38	73.45	0.80	47.06	52.45
SMOTE–SVM	0.45	63.58	0.70	49.02
SMOTE–MLP	0.33	61.44	0.86	55.88
SMOTE–RF	0.33	62.01	0.86	57.84
Subsample–NB	0.25	78.88	0.73	32.00	37.00
Subsample–SVM	0.52	84.82	0.60	32.00
Subsample–MLP	0.42	79.08	0.71	40.00
Subsample–RF	0.43	77.07	0.68	44.00