Hybrid System of PCA, Rough Sets and Neural Networks for Dimensionality Reduction and Classification in Human Face Recognition
Feature selection is the problem of choosing a small subset of features that is necessary and sufficient to describe target concept. The importance of feature selection is due to the potential for speeding up the processes of both concept learning reducing the cost of classification, and improving the quality of classification. In this paper, a face recognition system based on rough set theory and neural networks is developed in this study, which PCA (Principal Component Analysis) approach has been used to reduce Feature vector, which is basically transformation of space. For selection of feature the authors have used the concept of reduct and core from rough set theory. Classification of face was realized using Learning Vector Quantization (LVQ) neural network.