Neural Network Optimal Design Based on Orthogonal Experiment Method

In this paper, using orthogonal experiment method, taking three layers feed forward neural network as a example, problem of choosing training samples, weights and bias, training parameter of neural network is analyzed and studied. And on the basis of it, a neural network optimal design method based on orthogonal experiment method is carried out. The research result proved that using orthogonal experiment method in the conform of neural network model, an effective result can be obtained, and it is an intuitionistic, easy, objective and scientific method.