Expansion Type Functional Neuron Network Model and Its Parameters to Directly Determine the Method

In this paper, functional neuron structure will be expanded and deformed in the functional networks, then the expansion type functional neuron network model is obtained, and the pinv Moore-Penrose of matrix is employed, the optimal value of the functional parameters in the functional networks can be rapidly and accurately determined directly. Finally simulation experiments show that the proposed method is feasible and effective, can obtain higher approximation accuracy.