Table 1 The experimental results for the seven classifiers
Classifier | Training cohort | Test cohort | ||||||
|---|---|---|---|---|---|---|---|---|
ACC (%) | AUC | SPE (%) | NPV (%) | ACC (%) | AUC | SPE (%) | NPV (%) | |
Optimizable Tree | 77.50 | 0.61 | 96.80 | 77.20 | 70.30 | 0.49 | 96.20 | 71.40 |
Logistic Regression | 64.00 | 0.65 | 73.00 | 75.40 | 67.60 | 0.60 | 84.60 | 73.30 |
Kernel Naive Bayes | 73.00 | 0.69 | 90.50 | 76.00 | 62.20 | 0.63 | 84.60 | 68.80 |
Linear SVM | 85.40 | 0.85 | 98.40 | 83.80 | 75.70 | 0.77 | 96.20 | 75.80 |
SVM Kernel | 73.00 | 0.64 | 96.80 | 73.50 | 73.00 | 0.75 | 100.00 | 72.20 |
Optimizable Ensembles Learning | 77.50 | 0.73 | 88.90 | 81.20 | 75.70 | 0.80 | 88.50 | 79.30 |
Narrow Neural Network | 76.40 | 0.76 | 88.90 | 80.00 | 70.30 | 0.68 | 88.50 | 74.20 |