Research on the Application of RBF NN on Enterprise Early-warning

Abstract

Enterprise early-warning management is a new mode in the modern enterprise management. The introduction of Neural Network (NN) into enterprise management is one of the newly development orientations of artificial intelligence. The Radial Basis Function (RBF) NN was introduced into enterprise early-warning management. The method of compound index was applied to deal with the indexes. And a dynamic nearest neighbor-clustering algorithm was putted forward to computer the hidden nodes’ number and center value, which overcome the dependence of most neighbor algorithm on parameters. Using the monitoring indexes based on the compound index method, the mathematics model was constructed. Simulation result that adopted RBF NN indicates that the method has good effect, which also proves the validity of this method.