Research Article
Wavelet Scattering Transform for ECG Beat Classification
Table 9
Summary of classification results achieved by all the methods in this paper.
| Feature extraction | Classifier | TP | TN | FP | FN | ACC (%) | PPV (%) | SEN (%) | SPEC (%) |
| WSN | NN | 261101 | 88681 | 1319 | 8899 | 97.2 | 99.5 | 96.7 | 98.5 | WSN+PCA | NN | 263152 | 83508 | 6492 | 6848 | 96.3 | 97.6 | 97.5 | 92.3 | PNN | 268076 | 86738 | 3262 | 1924 | 98.6 | 98.8 | 99.3 | 96.4 | KNN | 266854 | 87816 | 2184 | 3146 | 98.5 | 99.2 | 98.8 | 97.6 | WSN+the 3th, 4th, and 5th time window | NN | 264158 | 88146 | 1854 | 5842 | 97.9 | 99.3 | 97.8 | 97.9 | WSN+the 4th time window | NN | 265091 | 87999 | 2001 | 4909 | 98.1 | 99.3 | 98.2 | 97.8 | PNN | 269828 | 86415 | 3585 | 172 | 99.0 | 98.7 | 99.9 | 96.0 | KNN | 268611 | 88915 | 1085 | 1389 | 99.3 | 99.6 | 99.5 | 98.8 |
|
|
TP: true positive; TN: true negative; FP: false positive; FN: false negative; WSN: wavelet scattering network.
|