企業財務危機預測,在過去一直是個在各公司經營管理者所熱門探討的話題,也是研究學者熱門的研究探討主題。若能從歷史資料中將一系列公司的財務屬性與財務危機以人工智慧資料探勘的方式進行分析而非如傳統僅以統計分析, 更能找出關鍵因子與門檻值,因此本研究的研究樣本取自1996年至2007年間於中國上海與深圳掛牌共884上市公司,其中268家破產公司,另616家為正常公司,並使用了決策樹方法以70%為期望的正確率,程式會自動從樣本裡隨機抽出30筆有效資料,以(a)完全健康的公司 (b)瀕臨倒閉的公司為兩個研究案例, 分別繪出決策樹後,關鍵因子和門檻值均被合理的有效計算。
The prediction of financial on enterprise in the past has drawn a great deal of attraction on studies for researchers and the manager of the enterprise. If we can use Artificial Intelligence or Data-Mining Techniques to analyze the financial attributes of a series of given companies which will render a better results rather than only used traditional statistical analysis. The core factors and thresholds can be computed through Decision Tree analysis. Hence, we adopt the data from 1996 to 2007 on China to make a practical analysis in which there are 884 companies(616 normal companies vs. 268 bankrupt companies). We also set the accuracy rate as 70% for Decision Tree analysis. The entire study has two study cases as: (a)completely healthy company and; (b)bankruptcy company as case studies. The Decision Tree is drawn rationally and core factor/thresholds are effectively computed.