Skip to main content

Table 2 Average G-mean for three cases

From: LVQ-SMOTE – Learning Vector Quantization based Synthetic Minority Over–sampling Technique for biomedical data

  G-mean  
Classification algorithm Nothing : base line AdaboostM1 SMOTE LVQ-SMOTE
NaiveBayes 76.25% 77.34% 78.54% 78.94%
Logistic Tree 72.88% 74.21% 81.21% 83.64%
Neural Network 75.24% 79.62% 80.44% 80.24%
SVM 72.65% 73.31% 80.92% 83.22%
RandomForest 75.34% 78.96% 79.47% 80.68%
OLVQ3 75.76% 74.35% 80.88% 82.55%
  1. Nothing represents that all the datasets were remained as the class imbalanced problem. In the case of SMOTE and LVQ-SMOTE, the minority samples were increased up to the number of the majority samples.