Research On Credict Risk Evaluation Of Tourism Enterprises Based On LSSVM

Since 2013,China's tourism industry keeps a stable and faster growth and becomes a force to reckon with in the country's GDP,playing a positive role in the continued and healthy growth of the national economy.However,the rapid development of tourism triggered the imbalance between supply and demand.The shortage of funds also became a limitation of the development of tourism.For investors,as the risk of tourism investment increases,more and more bad cases of tourism investment appears.Investors also transform the way of investment decision from blindness to rational.Rational investment requires investors to evaluate the investment risks of tourism enterprises accurately.The credit risk of the enterprise is a core indicator of investment risk evaluation.For tourism enterprises,the criteria for corporate credit risk evaluation are not clear,and internal control over credit risk cannot be achieved.For the entire tourism industry,the general credit risk evaluation method is missing.Based on the existing research of credit risk evaluation,this paper considers the imperfect and unscientific problems in the credit risk evaluation of tourism enterprises.Considering the difficulties in financing of tourism enterprises,this paper summarizes and improves the credit risk evaluation index system of tourism enterprises.The index data obtained in this paper are analyzed by Principal Component Analysis with MATLAB software.According to the Principal Component Analysis,each principal component is substituted into the Least Square Support Vector Machine(LSSVM)model,and then the radial basis function is selected as the kernel function in the LSSVM evaluation model.Taking several tourism enterprises as examples to conduct empirical research,the operability of the credit risk evaluation model is proved,and compared with the results of Logistic regression model,ultimately the accuracy of the LSSVM credit evaluation model is higher.Finally,it provides feasible suggestions for investors to invest in the tourism industry,for tourism enterprises to improve their credit rating,and for tourism managers to better realize supervision.The conclusions of this paper are mainly three points: the credit evaluation index system of tourism enterprises constructed in this paper is effective and meaningful;applying the LSSVM method to the evaluation of credit risk of tourism enterprises is a new attempt,and its model prediction results are good;the LSSVM model is more suitable for tourism enterprise credit risk evaluation than the Logistic regression model in this paper.