A Novel Ensemble Method for Imbalanced Data Learning: Bagging of Extrapolation-SMOTE SVM

Table 1

The process of bootstrapping. The symbol marks the data which is sampled in the turn while ○ marks the data which is not sampled. Specifically, sets that are not sampled in each turn can be used as validation set for parameters’ tuning.

Training dataset
1		○
2	○
3	○		○
	○	○	○
		○	○
			○		○