Table 7 Classification accuracy versus the dataset's characteristics
Datasets | Number of classes | Dimension | Number of samples | Naïve Bayes | k-NN | SVM | |
|---|---|---|---|---|---|---|---|
Balance scale | original | 3 | 4 | 625 | 86.70% | 88.46% | 90.54% |
+3D | 7 | 88.14% | 92.63% | 90.87% | |||
Tic-Tac-Toe Endgame | original | 2 | 9 | 958 | 71.06% | 81.84% | 91.01% |
+3D | 12 | 78.16% | 85.39% | 93.10% | |||
German | original | 2 | 20 | 1000 | 72.97% | 69.00% | 69.97% |
+3D | 23 | 73.07% | 68.80% | 69.97% | |||
Ionosphere | original | 2 | 34 | 351 | 81.71% | 86.29% | 92.57% |
+3D | 37 | 80.86% | 90.29% | 93.14% | |||
Teaching assistant evaluation | original | 3 | 5 | 151 | 52% | 64.00% | 62% |
+3D | 8 | 53.33% | 70.67% | 63.33% |