Machine-Learning Approach to Optimize SMOTE Ratio in Class Imbalance Dataset for Intrusion Detection

Table 12

Results of the SVR model (10-fold cross-validation).

Measures		Values
Correlation coefficient		0.760
Mean absolute error		0.005
Root mean squared error		0.006
Relative absolute error		60.9%
Root relative squared error		64.6%
Standard deviation of SMOTE ratios	U2R	292.071
	R2L	158.932
	Probe	58.276