Research On Financial Early Warning Based On RBF Neural Network

Abstract:The principle of neural network is simulation of human brain thinking, neural network structure is composed of neuron and link arcs elements. During neural network calculation process, different neurons can process input data according to specific rules. Through data transfer between several layers of neurons, it can reach the goal of function approach by any precision. The use of neural network in financial early warning analysis can predict enterprise financial states by analysis of financial related data.RBF neural network is a kind of neural network structure that developed in recent years. This network work structure can improve the efficiency of network training by phased training technology for neural network. Principal component analysis is a kind of analysis of the data processing algorithm, through the principal component analysis, retains most of the variance in the data, in order to maximize reflected the differences of data. In order to attain the goal of financial early warning for different enterprise, prediction model that based on RBP neural network and principle component analysis algorithm is designed in this paper, and the model is used in enterprise financial early warning process. In this prediction model, principle component analysis technology is used for the pre-process of financial related dates. It retains the major variance items and ignores the minor factors. The data that processed by principle component analysis algorithm is used for RBF neural network input item in order to reduce input item quantity and enhance neural network process efficiency. The experiment result show that the financial early warning model established in this paper can enhance process efficiency and prediction precision in contrast with traditional RBF neural network model. Requirement of financial information system is analyzed, and the financial early warning model established in this paper is used ad core algorithm in this system. The financial early warning system is developed, and the use of information system enhanced the automation level for financial early warning.Research in this paper established financial early warning algorithm and information for the business of enterprise financial, and provided technical support for financial work.