Research Article
Disease Classification and Biomarker Discovery Using ECG Data
Table 5
Classification results of the different methods on the test set of cases.
| Data set | Method | Sensitivity | Specificity | Accuracy |
| Mean | SDA | 82.29% | 73.08% | 80.33% | SVM | 85.57% | 61.54% | 80.49% | LLR | 92.71% | 34.61% | 80.33% |
| Quantile | SDA | 89.58% | 73.04% | 86.66% | SVM | 86.6% | 73.07% | 83.74% | LLR | 86.46% | 69.23% | 82.79% |
| Mean + PCA | SDA | 87.5% | 73.08% | 84.43% | SVM | 89.7% | 50% | 81.3% | LLR | 89.58% | 38.46% | 78.69% |
| Quantile + PCA | SDA | 89.68% | 84.62% | 88.52% | SVM | 89.68% | 76.92% | 86.99% | LLR | 94.79% | 53.85% | 86.70% |
|
|