An early warning model for risk management of securities based on the error inverse propagation neural network

This paper has carried on the tentative discussion to the early warning model for risk management under the separation of the sectors of investment bank. The characteristic of the investment bank is analysed at first, then discern the risk of the investment bank to classify. The index system of early warning model is proposed. Propagating against error neural network of neural network theory is introduced. This network is utilised to build early warning model and realize the corresponding algorithm by the tool MATLAB. The theory and realistic meaning of this paper is to promote the development of investment bank theory of our country, opening up the research range of the financial theory, especially there are important theory meaning and applying value to the setting-up of the early warning system of risk of the securities business.