Learning Algorithm And Its Application In The Network Of Multi Layer Radial Basis Function Networks Using Generalized Inverse Matrix

In this paper, the generalized inverse matrix is applied to the genetic algorithm of multi- layer radial basis function network, and the basic search method and genetic algorithm are used in the complex multi- layer radial basis function network.Computer experiments show that the accuracy of the approximation is improved by 1to 2 orders of magnitude by using this method. Further, the improved algorithm is used to predict chaotic time series, and the prediction step can be greatly improved. At the same time, the numerical solution of the boundary value problem of the partial differential equation is also improved by 1 to 2 orders of magnitude. Finally, as an application example, the genetic algorithm for the composite multilayer radial basis function network using the generalized inverse matrix is used to solve the option pricing equation, and the numerical solution with higher accuracy is obtained.