Table 8. Accuracy results of repeated double cross validation of the Random Forest and Linear Discriminant Analysis (LDA) models built to classify patients with prostate cancer and cancer-free controls based on blood PSA levels and urinary VOCs.
Repeated double cross validation (%) | ||||||||||||
Accuracy | Sensitivity | Specificity | ||||||||||
Model | Mean | Min | Median | Max | Mean | Min | Median | Max | Mean | Min | Median | Max |
Random Forest | 70.88 | 52.94 | 70.59 | 82.86 | 80.16 | 60.00 | 80.00 | 100.00 | 58.23 | 28.57 | 57.14 | 85.71 |
LDA | 65.09 | 47.06 | 64.71 | 80.00 | 75.56 | 45.00 | 75.00 | 100.00 | 50.80 | 14.29 | 50.00 | 85.71 |
Monte-Carlo repeated double cross validation (%) | ||||||||||||
Accuracy | Sensitivity | Specificity | ||||||||||
Model | Mean | Min | Median | Max | Mean | Min | Median | Max | Mean | Min | Median | Max |
Random Forest | 50.52 | 26.47 | 50.00 | 73.53 | 64.09 | 35.71 | 64.29 | 94.74 | 36.72 | 5.88 | 37.50 | 64.29 |
LDA | 49.89 | 32.35 | 50.00 | 72.73 | 64.70 | 38.46 | 65.00 | 90.00 | 34.70 | 0.00 | 33.33 | 81.82 |