Parameter selection of support vector machine based on chaotic particle swarm optimization algorithm

Support vector machine (SVM) are new methods based on statistical learning theory. Training SVM can be formulated as a quadratic programming problem. The parameter selection of SVM should to be done before resolving the QP problem. Particle swarm optimization (POS) algorithm was adopted to select parameters of SVM. To improve its global search ability, POS algorithm was modified by virtue of chaotic motion with sensitive dependence on initial conditions and ergodicity. It is shown by simulation that the chaotic POS algorithm can derive a set of optimal parameters.