Table 6 Comparison of classification accuracies obtained using different distance metrics
Datasets | Distance metrics | ||||
|---|---|---|---|---|---|
Original | Euclidean (+2D) | Chi-square 1 (+2D) | Chi-square 2 (+2D) | Mahalanobis (+2D) | |
Abalone | 20.37% | 50.95% | 48.17% | 56.26% | N/A* |
Balance scale | 58.24% | 64.64% | 85.12% | 78.08% | 76.16% |
Corel | 16.63% | 5.45% | 3.5% | 1.86% | N/A* |
German | 61.3% | 99.9% | 84.5% | 79.8% | 61.3% |
Hayes-Roth | 37.12% | 68.18% | 50.76% | 43.94% | 41.67% |
Ionosphere | 86.61% | 84.05% | 86.61% | 71.79% | N/A* |
Iris | 96% | 98% | 95.33% | 95.33% | 94% |
Teaching assistant evaluation | 58.94% | 66.23% | 64.9% | 65.56% | 64.9% |
Tic-Tac-Toe Endgame | 22.55% | 99.58% | 86.22% | 86.22% | 86.64% |