Facial expression recognition approach based on least squares support vector machine with improved particle swarm optimization algorithm

The problem in parameter selection of least squares support vector machine (LS-SVM) restricts the development of LS-SVM, In order to choose the optimal parameters of LS-SVM automatically, we proposed an improved particle swarm optimization (PSO) algorithm which can not only increase the convergent speed but also improve the overall searching ability of the algorithm. The improved PSO algorithm can increases the ability of avoiding local optimum effectively. We use the improved PSO algorithm to choose the optimal parameters of LS-SVM automatically in facial expression recognition system. The experimental results show that the proposed LS-SVM method with improved PSO is superior to BP network, traditional SVM, and PSO-SVM. We can achieve higher recognition accuracy and higher velocity of convergence by using the proposed method.