Comparative study of Machine Learning, Deep Learning and Bayesian Models on the ORL Faces Dataset for Effectual Face Recognition

This study compares face recognition performance between machine learning, deep learning, and probabilistic reasoning models. On the ORL dataset from Kaggle, the Machine Learning (ML) model (support vector machine (SVM)) achieved a recognition accuracy of 96.25.%, The Deep Learning (DL) model convolutional neural network (CNN) achieved a recognition accuracy of 99.99.% and a Probabilistic (PR) model Bayesian network attained an accuracy of 90.%. This analysis emphasizes CNN’s superior accuracy in face recognition, especially on large-scale datasets, while also illuminating the comparatively limited effectiveness of Bayesian Networks with PCA and SVM for dimensionality reduction and classification.