Table 1 Final output of the CPI parameters obtained after twenty iterations of linear logistic regression, ANN, linear- and RBF-SVM models
From: Non-invasive detection of fasting blood glucose level via electrochemical measurement of saliva
S. no. | Machine learning technique | Computational parameters | |||
|---|---|---|---|---|---|
Accuracy | Precision | Recall | F1 score | ||
1 | Linear logistic regression | 75.86 ± 2.3 | 76.76 ± 3.8 | 75.48 ± 5.4 | 75.71 ± 2.6 |
2 | ANN | 80.7 ± 2.1 | 81.2 ± 1.7 | 79.3 ± 3.4 | 80.2 ± 2.2 |
3 | Linear-SVM | 77.93 ± 2.7 | 77.59 ± 3.5 | 79.43 ± 4.7 | 78.11 ± 2.7 |
4 | RBF-SVM | 84.09 ± 2.8 | 83.75 ± 3.3 | 84.92 ± 4.5 | 84.06 ± 2.9 |