Optimized features selection for gender classification using optimization algorithms

Optimized feature selection is an important task in gender classification. The optimized features not only reduce the dimensions, but also reduce the error rate. In this paper, we have proposed a technique for the extraction of facial features using both appearance-based and geometric-based feature extraction methods. The extracted features are then optimized using particle swarm optimization (PSO) and the bee algorithm. The geometric-based features are optimized by PSO with ensemble classifier optimization by the genetic algorithm. Using this approach, we have obtained promising results in terms of the classification error rate and computation time minimization. Moreover, our optimized feature-based method is robust to illumination, noise, and occlusion changes.