Table 4 Comparison of accuracy of 5-fold cross validation based on structured data models
From: Clinical assistant decision-making model of tuberculosis based on electronic health records
| Â | 1-Fold | 2-Fold | 3-Fold | 4-Fold | 5-Fold | Mean | STD |
|---|---|---|---|---|---|---|---|
SS-PTDM | 0.9505 (0.9456, 0.9554) | 0.9471 (0.9421, 0.9521) | 0.9569 (0.9524, 0.9614) | 0.9498 (0.9449, 0.9547) | 0.9448 (0.9397, 0.9499) | 0.9498 (0.9458, 0.9538) | 0.0046 |
XGBoost | 0.9337 (0.9281, 0.9393) | 0.9301 (0.9244, 0.9358) | 0.9305 (0.9248, 0.9362) | 0.9285 (0.9227, 0.9343) | 0.9298 (0.9241, 0.9355) | 0.9305 (0.9288, 0.9322) | 0.0019 |
Random Forest | 0.8952 (0.8883, 0.9021) | 0.8971 (0.8903, 0.9039) | 0.8895 (0.8825, 0.8965) | 0.9086 (0.9021, 0.9151) | 0.9074 (0.9009, 0.9139) | 0.8996 (0.8924, 0.9068) | 0.0082 |
SVM | 0.9376 (0.9322, 0.943) | 0.9325 (0.9269, 0.9381) | 0.9331 (0.9275, 0.9387) | 0.9282 (0.9224, 0.934) | 0.9242 (0.9183, 0.9301) | 0.9311 (0.9266, 0.9356) | 0.0051 |