The Research On The Method Of The Credit Risk Forecasting About Listed Tourism Companies

The credit risk is one of the most important risks in the enterprise operation process, and is indispensable as a piece of research content of the enterprise management field study. Chinese tourism industry started late, but as a sunrise industry, tourism has very good prospects for development. Many tourism companies are developed, at the same time, lots of problems appear. For example, the governance structure is unreasonable, peer competition, the internal management is confusion, and credit crisis. Especially the credit risk not only makes the enterprise suffered huge losses, also influences the development of enterprises and the tourism. Therefore, the method of effectively predicting enterprise credit risk becomes the interested problem to the enterprise and scholars.In this way, the enterprise can take measures to control and prevent the credit risk timely.At first, selecting28companies listed tourism companies in the Shanghai and Shenzhen stocking exchange listing as the credit risk prediction of the enterprise samples. And the experiment used the whole sample to analyze the sample. With the selecting principle of variable index, the15financial ratios were chosen as the samples variables. Meanwhile, collecting the enterprise dates form2000to2010as the data sets. After the pretreatment of the data, describing the data sets,and analyses the mean value and the standard deviation of the data. At the same time, the normal sexual inspection proved the sample applicability of the experiment to study the credit risk prediction method. In order to eliminate the influence of different order and magnitude, the standardization of data processing was used.Then two methods were adopted for empirical research. The first method was the processing of the data set with four different methods. The methods were standardization, Random-sampling algorithm, manifold learning, combined with Random-sampling algorithm and manifold learning. Through comparison and analysis of the three models, such as MDA,Logit and Probit, obtained the concludes that the prediction accuracy was effective. With the way of combining the Random-sampling algorithm to balance the data set and the study of manifold ISOMAP and LLE algorithm to dimension reduction processing method. Also finding out the best prediction model which was Logit model. And the test proved the method of data processing and the combination of the Logit model can take the best forecasting accurate. Although we find the accurate forecast of ST year is high with the closer data. Another method is combining with factor analysis and the Logit model, and in this way getting good simulation model effect, achieving the good prediction accuracy. According of the index, concrete analysis of our listed tourism companies credit risk prediction effect was obtained, and concluded that the enterprise in the business process should pay attention to the financial index and its fluctuations.This paper discusses the credit risk prediction method study of the listed tourism companies. This test can carry out the credit risk prediction for the enterprise to provide technical support, also abounded the prediction of the experiment method. IT has certain theoretical and practical significance.