Fig 1.
Relationship between population, sample and synthetic data [10].
Fig 2.
The GSDOT data generation mechanism when positive class samples are generated.
The process is repeated for the negative class.
Table 1.
Description of the datasets.
Fig 3.
Visualization of the experimental procedure.
Table 2.
Results for mean cross validation scores of all methods.
Table 3.
Results for percentage difference between GSDOT and NONE.
Table 4.
Results for mean rankings of all methods.
Table 5.
Results for Friedman test.
Table 6.
The p-values of the Holm’s test.