Session:

GSW - Graduate Student Workshop

Title:

Differential Evolution Algorithm as a Tool of Training Radial Basis Function Networks

Authors:

Junhong Liu
Jouni Lampinen

Abstract:

The Differential Evolution (DE) algorithm is a floating-point encoded evolutionary algorithm for global optimization. It has been demonstrated to be an efficient, effective, and robust optimization method especially for problems containing continuous variables. This paper concerns applying DE to training the radial basis function (RBF) networks. It is demonstrated by training networks to approximate three nonlinear functions. The Euclidean distance from the desired outputs to the actual network outputs is applied as the objective function to be minimized. The process converges effectively. The results show that DE is a potential way to train Gaussian RBF networks.