Research Article
Usage of Probabilistic and General Regression Neural Network for Early Detection and Prevention of Oral Cancer
Table 5
Performance of PNN/GRNN model for predicting survivability of oral cancer patients.
| Performance estimation parameter | Performance |
| Positive/negative ratio | 0.674 | Accuracy | 69.95% | True positive (TP) | 36.68% | True negative (TN) | 33.27% | False positive (FP) | 26.44% | False negative (FN) | 3.61% | Sensitivity | 91.04% | Specificity | 55.72% | Geometric mean of sensitivity-specificity | 71.22% | Positive predictive value (PPV) | 58.11% | Negative predictive value (NPV) | 90.21% | Geometric mean of PPV and NPV | 72.41% | Precision | 58.11% | Recall | 91.04% | -measure | 0.709 | Area under ROC curve | 0.7491 |
|
|