| 摘要: | 本研究應用資料探勘方法之類神經網路、決策樹及支援向量機，作兩階段的篩選及建模分析。研究樣本為2006年至2015年發生財務危機公司。以同產業且資產總額相當，將財務危機公司與財務正常公司，並以 1：3比例配對。本研究共分為二個階段：第一階段應用類神經網路、決策樹QUEST和決策樹CHAID進行篩選變數。第二階段應用類神經網路、決策樹CART、決策樹C5.0和支援向量機，建立財務危機預測模型。本研究結果顯示，在第二階段建立模型中皆以決策樹C5.0的分類表現最佳。其中實證結果顯示，應用類神經網路進行變數篩選，再搭配決策樹C5.0建立模型，結果較佳，準確率為84.35%。對預測企業財務危機具有較佳的預測能力。
The Neural Network, Decision Tree, and Support Vector Machine (SVM) are used in this study, for a two-stage model. The period is from 2006 to 2015. In the first stage, Neural Networks, Decision Tree QUEST and Decision Tree CHAID are applied for screening variables. In the second stage, Neural Networks, Decision Tree CART, Decision tree C5.0 and Support Vector Machine are applied to establish the prediction model for financial distress. The results show that Neural Network (the first stage) and Decision Tree C5.0 (the second stage) has the best performance 84.35%. This kind of method proves to be more effective in predicting the financial distress of companies. |
