Research And Application Of Combinatorial Method For The Financial Risk Prediction

Under the condition of market economic, marketing competition is gradually vigorous, Survival in the harsh business environment, low efficiency, innovation ability, weak management capacity of enterprises will be eliminated.In recent years, enterprises which fall into crisis because of financialrisk is gradually vigorous. So how to make a efficacious model of finance risk prediction and valuation.This not only is the key point which the listed company pays attention to, but also become urgent need for various benefit correlations. The company which falling into the financial distress is a process from the common to the worse, and result in bankruptcy at last.This shows the auspices of the financial distress can be forecasted, the related research in home and abroad prove. This paper on the status and characteristics of the company, using finance crisis management theories and knowledge of accountant build a combination forecasting and ant colony algorithm model, analyses the finance management of listedcompanies and gets the solution.In order to improve the prediction accuracy of financial forecasting model, this paper presents the combination of ant colony algorithm based model for financial forecasting model. Main tasks are as follows:(1) Model preparationBefore modeling, the use of principal component analysis to reduce the dimensions of the financial indicators processed to extract the main ingredients.(2) Create single predictive modelsBased on characteristics of the financial crisis choose conditional logistic regression forecasting model?multiple linear regression forecasting model and BP network forecasting model as single forecasting model. After that based on financial data of 2008 to 2009, analyze empirical of three individual models.(3) Create a combination modelThe author improves ant colony algorithm to solve the problem of weightedaverage coefficients.This article formulated objective function according to least squares criterion to solve the weighted average, coefficientsand select the single eforecasting model by the weighted average coefficients.We eliminate the single forecasting model with smallest weighted averaged coefficient.Then the filtration of single forecasting model is qualitative and quantitative.(4) Designing for financial risk prediction subsystemAgainst to characteristics of financial risk prediction, and in considering the corporate information, financial data, financial indicators and other factors, design and development of the corresponding filter module and the prediction model. Demonstrate by an example the system to complete the function.(5) Empirical researchThe empirical study is by the law of large number of samolessummary of historical data, it is different from the perspective of empirical analysis of financial distress prediction. This result provedthe practicality of a better and higher prediction accuracy.