| Summary: | 碩士 === 國立彰化師範大學 === 企業管理學系 國際企業經營管理(IMBA) === 104 === The rapid changes of the economic environment resulting in an increase of the possibility of business failure and different kinds of reasons in the past two decades; More recently, the increase of exchange rate volatility jeopardize the healthiness of Taiwanese firms which highly depends on the international trade. The study intends to examine whether an inclusion of the foreign exchange risk variable enhance the accuracy of the early warning model.
The study utilize the TSE listed companies which encountered the financial crises as hazard firms and selected the normal counterparts by 1:2 ratio over the period from 2013 to September 2014. The empirical implementation proceeds as a 2-step approach. First of all, the study will extract predict variables by GA+SVM; secondly, apply those selected predict variables to formulate the early warning model including SVM and neural network model and logistic regression model framework and the results will be validated in terms of classification ratio, type error and prediction error, RMSE.
The empirical results show that an inclusion of the exchange rate risk increase the accuracy of GA-SVM based prediction model. Moreover, the alternative GA-SVM model outperforms the others in terms of both accuracy and prediction error.
|
|---|
