Cauchy mutation for decision-making variable of Gaussian particle swarm optimization applied to parameters selection of SVM

Due to the slow convergence of Gaussian particle swarm algorithm (GPSO) during parameters selection of support vector machine (SVM), this paper proposes a novel PSO with hybrid mutation strategy. Since random number generated from Cauchy distribution has better convergence characteristic than ones from Gaussian distribution during mutation strategy. Cauchy mutation is applied to amend the decision-making variable of Gaussian PSO. The adaptive mutation based on the fitness function value and the iterative variable is also applied to inertia weight of PSO. The results of application in parameter selection of support vector machine show the proposed GPSO with Cauchy mutation strategy is feasible and effective, and the comparison between the method proposed in this paper and other ones is also given, which proves this method is better than Gaussian PSO.