Face Classification Using Stacking Ensemble Learning

Developing a high-level automated security system for identification/authentication has long been an active research subject in practically all fields. Traditional security systems utilize a key or a password to secure a process/product, whereas biometric security systems use a person’s physical or behavioral attributes. Face patterns play a significant part in several potential recognition/authentication applications because of their uniqueness, universality, dependability, and stability. This study provides a face identification system that starts with eye detection and ends with face image recognition. AN ensemble stacking approach for face classification is presented in this study that combines various Machine learning (ML) algorithms with and without Principal Component Analysis (PCA). The proposed system is trained and tested on a well-known publically available Olivetti Face Dataset. The dataset is split into 75% train and 25% test dataset which is also transformed using PCA. The results show that the accuracy of each base model having PCA data is on par or better than the one without PCA. The paper also proves that the performance of the new stacking ensemble system results in an average or better performance than the other base models. For Olivetti face dataset, the proposed Stacked Ensemble Learning model’s accuracy with PCA is 94%, which is on average than the accuracy of the other base models, including Neural Network (NN) (100%), Random Forest (92%), Support Vector Machine (SVM)(96%), Gaussian Naives Bayes(81%), K-Nearest Neighbour (KNN) (89%), and Logistic Regression (LR) (97%) respectively.