Comparative Analysis of Dimensionality Reduction Techniques and Machine Learning Algorithms for Face Mask Detection

Abstract

This paper presents a comprehensive study on the application of Principal Component Analysis (PCA) and Linear Discriminant Analysis (LDA) in conjunction with various machine learning classifiers for the purpose of facial landmark detection and classification. The study evaluates the performance of five different classifiers: K-Nearest Neighbors (KNN), Support Vector Machine (SVM), Decision Tree, Gaussian Naive Bayes (GaussianNB), and Random Forest, in terms of accuracy, precision, recall, and F1 score. Both PCA and LDA are used for dimensionality reduction to enhance classifier performance. The results demonstrate significant differences in the effectiveness of these dimensionality reduction techniques when combined with different classifiers. The analysis provides insights into the optimal combinations of dimensionality reduction techniques and classifiers for facial landmark detection tasks.