財務預警模型建立的目的在於提早發現潛在的危機,過去學者討論財務預警模型,多使用因素分析、集群分析等統計方法來縮減變數,再由單變量分析、多變量區別分析、類神經網路模型等方法,來建構財務預警模型,縮減變數是否會遺失了重要的財務比率訊息.由於先前的研究未將財務比率加以探討,因此本研究的採用logistic迴歸模式將財務比率分別建構財務危機預警模型,並找出明顯的區別值供財務使用者作為判斷的依據。
The purpose of building predictive financial models is that find latent crisis advance the data. The researcher before study the predictive financial models usually use the statistics method, as factors analysis, cluster analysis, etc. They use those methods to reduce variable. Then the researcher building predictive financial model by university regression model, multivariate regression model, and neural networks. Will reducing the factor lose important information? Because earlier study never only discuss the financial ratios, the study will build financial distress prediction model by using logistic model and find obvious basis number to decide.