Research Article
Emotion Modeling in Speech Signals: Discrete Wavelet Transform and Machine Learning Tools for Emotion Recognition System
Table 2
The results of different cross-validation techniques and comparison with machine learning classifiers, in emotion classification tasks.
| Type | Cross-validation | Holdout validation | Resubstituting validation | 5 | 10 | 20 | 10 | 20 | 40 |
| SVM | 94.20 | 94.77 | 95.89 | 94.34 | 94.34 | 93.90 | 100.00 | KNN | 94.76 | 95.33 | 95.89 | 94.34 | 95.28 | 93.43 | 100.00 | Efficient Logistic Regression | 88.78 | 89.72 | 89.91 | 88.68 | 90.57 | 89.67 | 89.72 | Naive Bayes | 94.01 | 94.21 | 94.21 | 92.45 | 95.28 | 94.37 | 97.94 | Ensemble | 94.76 | 94.95 | 95.33 | 96.23 | 93.40 | 95.31 | 100.00 | Neural Network | 93.83 | 95.51 | 95.33 | 96.23 | 95.28 | 93.43 | 100.00 |
|
|