以決策樹為基礎之支援向量機模型於信用評等之研究

Application of Decision tree-based Support Vector Machine model for Taiwan Corporate Credit Risk Index

陳益軒(YI SHIEN TAN)

指導教授 ： 白炳豐

國立暨南國際大學/管理學院/資訊管理學系/碩士(2013年)

https://doi.org/10.6837/NCNU.2013.00283

摘要

隨著資本主義的蓬勃發展之下，各大企業以及中小企業的產生，在快速發展的全球經濟之下，各行各業中企業會透過財務槓桿的運用與銀行之間作交換，進而獲取資金轉為投資。對於金融業來說，一但市場波動發生變化，會具有無可比擬的破壞力。所以有鑑於此，在眾多的銀行機構才需要一套具有公信力的評等機制；透過第三方所評選的評等機制，可以提供給銀行業對每家企業具有基本放款依據。在研究中主要採取決策樹為基礎之支援向量機來針對信用評等評選機制作分類，透過以決策樹為基礎的支援向量機(DT-SVM)、無向循環圖支援向量機(DAGSVM)、傳統一對一支援向量機(OVOSVM)、一對多支援向量機(OVASVM)對信用評等資料分類，藉由最佳化參數期望可以試著找尋出最佳分類模組。並透過對分類準確率作出比較，從中評選出適合之分類模型以供銀行機構做參考依據。最後在此研究中結果顯示在多類資料中，決策樹為基礎之支援向量機(DT-SVM)具有良好的分類以及效率。

關鍵字

信用評等 ； 支援向量機 ； 決策樹 ； 金融 ； TEJ信用風險指標TCRI

並列摘要

Below with the rapid development of capitalism, large corporations and small and medium enterprises appearing, in the rapid development of the global economy, all walks of life in the enterprise through the use of financial leverage and exchange between banks, thereby obtaining funds into investment. For the financial industry, there have an unparalleled destructive power when market volatility . So with this in mind, many banking institutions need a credible ratings system through mechanisms such as the selection of assessment by third parties, can be given to the banking sector to every enterprise has a basic loan basis. Mainly taken in the study of decision tree based on support vector machine for credit rating evaluation system for classification, through decision tree based on support vector machine (DT-SVM) on the credit rating data classification, through optimization of parameter expectations can try to find out the optimum classification parameters. Compared to the final classification accuracy, from which the selected fit the classification of model for banking institutions to make reference. Final results show many types of data in this study, based on decision tree SVM (DT-SVM) has a excellent classification, and efficiency.

並列關鍵字

Credit ratings ； SVM ； Decision-Tree ； finance ； TCRI

參考文獻

References

1. Angelini, E., Di Tollo, G., & Roli, A. (2008). A neural network approach for credit risk evaluation. The Quarterly Review of Economics and Finance, 48(4), 733-755.

2. Arun Kumar, M., & Gopal, M. (2010). A hybrid SVM based decision tree. Pattern Recognition, 43(12), 3977-3987.

3. Bennett, K., & Blue, J. A. (1998). A support vector machine approach to decision trees. Paper presented at the Neural Networks Proceedings, 1998. IEEE World Congress on Computational Intelligence. The 1998 IEEE International Joint Conference on.

4. Breiman, L. (1996). Bagging predictors. Machine learning, 24(2), 123-140.