WNN optimization design based on Artificial Fish-Swarm Algorithm

The problem such as parameters initialization and network structure determination is collectively referred to as the WNN optimization design. Aiming at the effect on the performance of WNN, an optimization design algorithm, which is based on Artificial Fish-Swarm Algorithm(AFSA), is proposed. The AFSA can synchronously determine the initial values of parameters and hidden layer nodes number in search space. The simulation results show it is an effective algorithm, which not only has higher accuracy and faster convergence rate but also can avoid the blindness of the WNN optimization design.