Application Research Of Artificial Neural Networks To Enterprise Economic Prediction And Early-warning System

This paper systematically summarized the research work in the field of economic prediction and early-warning system using artificial neural networks. Based on these works, this paper present the basic frame of enterprise early-warning system using ANN and expound each of the subsystems specifically.The following points are the main research works in this paper: (1) Using the ANN to predict the warning indexes is studied, and the feasibility is proved. (2) Construct the warning indexes system. Seven indexes that uncover the essence of business crisis are selected based on the definition of business crises. (3) The seventy companies financial report data are processed through Factor Analysis method. The formative training sample data are used to train the ANN model. (4) The early-warning system base on ANN is used to Simulate warning by the new data, and the warning results are satisfying.This paper preliminarily proves the feasibility of economic predicting the early-warning system using ANN. But some difficulties in the proposed methods and concrete practice need further research in the future.