Research Article
Predicting Metabolic Syndrome Using the Random Forest Method
Table 5
Decision rules extracted from one of twenty trees from the predictive model trained with Random Forest.
| | Frequency (%) | Error (%) | Condition | Prediction |
| | 42 | 0 | WC ≤ 79.75 and TG ≤ 150.5 | Non-MS | | 10.1 | 0 | FPG ≤ 99.5 and TG ≤ 149.5 and HDL-C > 49.15 and LDL-C ≤ 128.05 | Non-MS | | 9.6 | 0 | WC > 87.75 and systolic > 127 and TG > 149.5 | MS | | 3.3 | 0 | Sex = female and WC > 79.5 and systolic > 125 and FPG > 99.5 | MS | | 1.7 | 0 | WC > 87.5 and systolic > 128 and FPG > 99.5 | MS | | 1.7 | 0 | Sex = female and WC > 80 and TG > 150.5 and HDL-C ≤ 48.1 | MS | | 1.7 | 0 | WC ≤ 87.75 and systolic ≤ 127 and Hct > 44.77 | Non-MS | | 1.3 | 0 | Sex = female and WC > 79.5 and systolic > 129 and HDL-C ≤ 49.75 | MS | | 8.7 | 0 | FPG ≤ 99.5 and TG ≤ 149.5 and HDL-C > 49.15 | Non-MS | | 1.1 | 0 | FPG ≤ 99.5 and TG ≤ 149.5 and HDL-C > 38.7 and Hct > 44.1 | Non-MS | | 1.1 | 0 | Systolic > 125 and FPG > 99.5 and TG > 149.5 | MS | | 1.5 | 0 | WC ≤ 78.5 and HDL-C > 50.45 | Non-MS | | 1 | 0 | WC > 87.75 and FPG > 99.5 and TG > 148.5 | MS | | 1.5 | 1.1 | WC ≤ 87.5 and TG ≤ 150.5 and Hct > 42.305 | Non-MS | | 1.8 | 1.9 | FPG ≤ 99.5 and TG ≤ 149.5 and HDL-C > 39.6 | Non-MS | | 2.1 | 2.5 | Systolic ≤ 125 and diastolic ≤ 85 and TG ≤ 149.5 and HDL-C > 47.2 | Non-MS | | 2.5 | 2.9 | Sex = male and WC ≤ 87.75 and HDL-C > 39.95 | Non-MS | | 1 | 6.8 | Diastolic > 85 and TG > 149.5 | MS | | 1.6 | 6.5 | WC ≤ 87.75 and systolic ≤ 127 and FPG ≤ 99.5 | Non-MS | | 1.1 | 0 | TG > 150.5 and HDL-C ≤ 40.05 | MS |
|
|
