Research Article
Detection of Human Stress Using Optimized Feature Selection and Classification in ECG Signals
Table 3
Performance and comparison results of the proposed system with conventional works.
| References | Methodology | Experimental results in % | Precision | Recall | Accuracy | F1 score |
| Proposed | | 92.78 | 91.56 | 92.43 | 95.86 | [7] | FCM clustering | 87.65 | 86.32 | 87.39 | 89.39 | [8] | Convolutional neural networks | 86.23 | 85.56 | 90.19 | 91.50 | [9] | Long short-term memory (LSTM) network | 85.18 | 86.49 | 88.13 | 89.16 | [10] | Frequency analysis | 73.98 | 74.12 | 75 | 76.30 | [11] | Heart-rate variability (HRV) correlation analysis | 87.06 | 88.10 | 89 | 91 | [12] | Convolutional neural networks | 60.72 | 62.59 | 63.97 | 68.23 | [13] | Deep ECGNet | 80.35 | 81.79 | 82.7 | 85.26 | [14] | SVM and ANN | 88.26 | 88.79 | 89.21 | 89.96 | [15] | Minimum redundancy maximum relevance (mRMR) selection algorithm | 82.52 | 83.3 | 84.4 | 85.23 |
|
|