Research Article
An Expert System Based on Fisher Score and LS-SVM for Cardiac Arrhythmia Diagnosis
Table 4
Comparison of proposed expert system with the studies in the literature.
| Method | Performance criteria | Maximum classification accuracy |
| VFI5-GA [6] | 10-fold-cross-validation | 68% |
| DTI-HITON |
10-fold-cross-validation | 71.87% | NN-HITON [7] | 60.38% | KNN-HITON | 65.30% |
| OneR |
70% train-%30 test | 58.09% | J48 [8] | 74.26% | Naïve Bayes | 75.00% |
| AIRS-FWP [9] | 10-fold-cross-validation 80% train-%20 test | 76.20% 80.71% |
| RF-CBFS [10] RF-CBFS with random sampling | 10-fold-cross-validation | 76.30% 90.00% |
| Proposed expert system (FS-LS-SVM) | 10-fold-cross-validation | 82.09% |
|
|