Table 5 The T2DM classification models performance comparison
Model | Accuracy | Kappa | AUC | F-Measure | Sensitivity | Specificity |
|---|---|---|---|---|---|---|
Decision Tree | 89.38% +/- 5.93% | 0.787 +/- 0.119 | 0.852 +/- 0.155 | 88.51% +/- 6.83% | 83.75% +/- 10.29% | 95.00% +/- 6.45% |
Gradient Boosted Trees | 88.75% +/- 5.74% | 0.775 +/- 0.115 | 0.953 +/- 0.050 | 88.46% +/- 6.17% | 87.50% +/- 10.21% | 90.00% +/- 7.91% |
Random Forest | 92.50% +/- 3.95% | 0.850 +/- 0.079 | 0.989 +/- 0.017 | 92.39% +/- 4.27% | 92.50% +/- 8.74% | 92.50% +/- 6.45% |
Naïve Bayes | 90.00% +/- 6.04% | 0.800 +/- 0.121 | 0.967 +/- 0.032 | 89.69% +/- 6.44% | 88.75% +/- 10.94% | 91.25% +/- 8.44% |
KNN | 81.88% +/- 12.31% | 0.637 +/- 0.246 | 0.874 +/- 0.129 | 80.89% +/- 12.07% | 76.25% +/- 12.43% | 87.50% +/- 15.59% |