| ========== BiLSTM Classification Report ========== | |
| precision recall f1-score support | |
| Enjoyment 0.6490 0.7296 0.6869 991 | |
| Fear 0.5580 0.4709 0.5108 327 | |
| Sadness 0.4580 0.4747 0.4662 356 | |
| Anger 0.6587 0.6748 0.6667 369 | |
| Other 0.6601 0.6733 0.6667 600 | |
| Disgust 0.4967 0.4488 0.4715 332 | |
| Surprise 0.4683 0.3620 0.4083 326 | |
| accuracy 0.5956 3301 | |
| macro avg 0.5641 0.5477 0.5539 3301 | |
| weighted avg 0.5893 0.5956 0.5905 3301 | |
| ========== Additional Metrics ========== | |
| Test Loss: 2.0363 | |
| Test Accuracy: 0.5956 | |
| Precision (Macro): 0.5641 | |
| Precision (Weighted): 0.5893 | |
| Recall (Macro): 0.5477 | |
| Recall (Weighted): 0.5956 | |
| F1-Score (Macro): 0.5539 | |
| F1-Score (Weighted): 0.5905 | |
| ========== Confusion Matrix ========== | |
| [[723 23 83 3 81 29 49] | |
| [ 38 154 26 72 10 14 13] | |
| [108 14 169 2 30 23 10] | |
| [ 13 42 12 249 14 29 10] | |
| [110 9 30 9 404 18 20] | |
| [ 32 25 26 30 38 149 32] | |
| [ 90 9 23 13 35 38 118]] |