A Comparative Analysis of K-Nearest Neighbors and Support Vector Machine for Human Emotion Detection Based on Voice Patterns to Improve Accuracy

The main goal of this research was to develop a model for classifying human emotions using a Novel K-Nearest Neighbors (KNN) algorithm and to compare its performance with a Support Vector Machine (SVM) algorithm. We used KNN with a sample size of <tex xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">$\mathbf{1 0}$</tex>, splitting the data into <tex xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">$\mathbf{7 0 \%}$</tex> for training and 30 % for testing. The study was conducted with specific parameters: alpha and beta values of 0.05 and 0.2, a 95 % confidence level, a G power of 80 %, and an enrollment ratio of 1. After applying both algorithms, we found that the Novel KNN algorithm achieved an accuracy of 87.5 %, which was higher than the SVM's accuracy of <tex xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">$\mathbf{8 5. 2 3 \%}$</tex>. The statistical analysis showed a significant difference between the two algorithms, with a pvalue of 0.04 (<tex xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">$\mathbf{p}; \mathbf{0. 0 5}$</tex>). This indicates that the KNN algorithm performed more effectively in classifying emotions compared to the SVM. Overall, the Novel KNN demonstrated better accuracy and efficiency in emotion classification.