Research Article
Decision Support System (DSS) for Fraud Detection in Health Insurance Claims Using Genetic Support Vector Machines (GSVMs)
Table 3
Summary performance metrics of SVM classifiers on samples sizes.
| Description | Kernels used | Data size | Average accuracy rate (%) | Sensitivity (%) | Specificity (%) |
| Linear | 100 | 71.43 | 60.00 | 77.78 | 300 | 72.73 | 84.21 | 0.00 | 500 | 91.80 | 97.78 | 75.00 | 750 | 84.42 | 95.00 | 47.06 | 1000 | 82.95 | 85.42 | 80.00 |
| Polynomial | 100 | 71.43 | 66.67 | 72.73 | 300 | 72.73 | 88.24 | 20.00 | 500 | 96.72 | 100.00 | 86.67 | 750 | 80.52 | 96.36 | 40.91 | 1000 | 84.71 | 83.67 | 86.11 |
| Radial basis function | 100 | 71.43 | 57.14 | 85.71 | 300 | 95.45 | 95.00 | 100.00 | 500 | 99.18 | 100.00 | 96.30 | 750 | 82.56 | 96.88 | 40.91 | 1000 | 90.91 | 100.00 | 82.98 |
|
|