SVM-based face recognition using genetic search for frequency-feature subset selection

Abstract

The goal of this paper is to study if there is a dependency between a selected feature vector at each generation of the genetic algorithm and the resulting fitness. In order to see the relation between these parameters, we first use Discrete Cosine Transforms (DCT) to transform each image as a feature vector (i.e., Frequency Feature Subset (FFS)). A Genetic Algorithm (GA) is then used to select a subset of features from the low-dimensional representation by removing certain DCT coefficients that do not seem to encode important information about recognition task. When using SVM, two problems are confronted: how to choose the optimal input feature subset for SVM, and how to set the best kernel parameters. Therefore, obtaining the optimal feature subset and SVM parameters must occur simultaneously. We present a genetic algorithm approach for feature selection and parameters optimization to solve this kind of problem. © 2008 Springer-Verlag.