Table 7 Comparison of the results obtained in two stages.
From: EEG based epileptic seizure detection using SVM fuzzy learning and metaheuristic optimization
Methodology | Performance |
|---|---|
Authors:42 Database: TUH Domain: time and frequency Classifier: Random Forest Dimensionality reduction: None Channels: 19 Features: 19 Attributes: 19 × 19 = 361 | Accuracy: 0.91 F1-score: 0.91 AUC: 0.95 |
Authors:43 Database: TUH Domain: time and frequency Classifier: SVM Dimensionality reduction: None Channels: 21 Features: 13 Attributes: 21 × 13 = 273 | Accuracy: 0.93 Precision: 0.94 Recall: 0.94 Specificity: 0.91 F1-score: 0.94 |
Authors:44 Database: TUH Domain: frequency Classifier: Logistic Regression Dimensionality reduction: PCA Channels: 21 Features: 5 Attributes: 7 | Precision: 0.73 Recall: 0.79 F1-score: 0.78 |
Authors:45 Database: TUH Domain: frequency Classifier: LightGBM Dimensionality reduction: None Channels: 21 Features: 96 Attributes: 2016 | F1-score: 0.84 |
Authors:46 Database: TUH Domain: frequency Classifier: CatBoost Dimensionality reduction: None Channels: 21 Features: 96 Attributes: 21 × 96 = 2016 | Accuracy: 0.88 Recall: 0.83 Specificity: 0.91 |
Authors:47 Database: CHB-MIT Domain: time and frequency Classifier: SVM Dimensionality reduction: None Channels: 22 Features: 18 Attributes: 22 × 18 = 396 | (Interictal; Ictal) Accuracy: 0.99;0.99 Precision: 1;0.98 Recall: 1;1 F1-score: 0.99;0.99 |
Authors:48 Database: CHB-MIT Domain: time and frequency Classifier: Random Forest Dimensionality reduction: None Channels: 20 Features: 8 Attributes: 1328 | Accuracy: 0.99 Recall: 0.99 Specificity: 0.83 |
Authors:49 Database: CHB-MIT Domain: frequency Classifier: KNN Dimensionality reduction: t-SNE Channels: 21 Features: 1 Attributes: 2 | Precision: 0.47 Recall: 0.81 F1-score: 0.56 |
Proposed Phase: 1 Database: CHB-MIT Domain: time and frequency Classifier: SVM-Fuzzy Dimensionality reduction: None Channels: 19 Features: 34 Attributes: 19 × 34 | Accuracy: 73.98 Precision: 70.51 Recall: 69.42 F1-score: 71.98 |
Proposed Phase: 2 Database: CHB-MIT Domain: time and frequency Classifier: SVM-Fuzzy Dimensionality reduction: Matrix Channels: 19 Features: 34 Attributes: 20 | Accuracy: 99.12 Precision: 97.98 Recall: 98.43 F1-score: 98.63 |
