Table 4 Multiclass classification performances of machine learning models in estimation of oral cancer survival time
From: Survival estimation of oral cancer using fuzzy deep learning
Models | Class | Precision | Recall (Sensitivity) | F1-score | Accuracy | Specificity | AUC of ROC curve |
|---|---|---|---|---|---|---|---|
Support Vector Machines | Class 0: 0 – 12 months | 0.53 | 0.26 | 0.35 | 0.74 | 0.77 | 0.59 |
Class 1: 13 – 24 months | 0.13 | 0.46 | 0.20 | 0.59 | 0.90 | 0.53 | |
Class 2: 25 – 36 months | 0.11 | 0.14 | 0.13 | 0.88 | 0.94 | 0.53 | |
Class 3: 37 – 48 months | 0.13 | 0.25 | 0.17 | 0.91 | 0.97 | 0.59 | |
Class 4: 49 – 60 months | 0.00 | 0.00 | 0.00 | 0.97 | 0.99 | 0.49 | |
Class 5: > 60 months | 0.62 | 0.35 | 0.45 | 0.55 | 0.52 | 0.56 | |
| Â | Â | Â | Â | Overall accuracy = 0.90 | Â | Â | |
Random Forest | Class 0: 0 – 12 months | 0.55 | 0.68 | 0.61 | 0.77 | 0.87 | 0.74 |
Class 1: 13 – 24 months | 0.29 | 0.31 | 0.30 | 0.84 | 0.91 | 0.61 | |
Class 2: 25 – 36 months | 0.00 | 0.00 | 0.00 | 0.92 | 0.94 | 0.49 | |
Class 3: 37 – 48 months | 0.00 | 0.00 | 0.00 | 0.95 | 0.96 | 0.49 | |
Class 4: 49 – 60 months | 0.00 | 0.00 | 0.00 | 0.99 | 0.99 | 0.50 | |
Class 5: > 60 months | 1.00 | 1.00 | 1.00 | 1.00 | 1.00 | 1.00 | |
| Â | Â | Â | Â | Overall accuracy = 0.91 | Â | Â |