A Novel Method for Face Recognition and Facial Expression Identification using PCA, Neural Network and Random Forest Classifier

The main aim of this paper is to analyze the method of Principal Component Analysis (PCA) and its performance when applied to face recognition. In previous work [10], they use Euclidian Distance for recognition. Here we use Neural Network and Random Forest Classifier which has much higher recognition accuracy than other methods. This algorithm creates a subspace (face space) where the faces in a database are represented using a reduced number of features called feature vectors and the classifier calculates the similarity score for performance evaluation which will provide improved results in terms of recognition accuracy.