銀行業為一財務槓桿程度極高的產業，在金融競爭環境日益強烈下，賺取而得的利差隨之越來越薄，致使收取的利息逐漸不足以沖抵呆帳損失。因此，授信良窳帶給銀行為利益或損失，端視各銀行對授信風險管理之重視程度而定。在風險管理領域中，良好的預警制度與明確的授信目標，將有助於銀行放款人員進行授信業務之評估參考。

　　本研究導入支撐向量機(Support Vector Machines, SVM)理論於財務金融運用上，將之與銀行授信決策結合，以財務比率為研究變數，經由因素分析法萃取重要變數，分析比較支撐向量機、類神經網路與Logit傳統統計方法三者在授信決策上之準確度。經由實證研究與分析發現：

一、變數選取：因素分析法能有效萃取重要因素以減低資料及模型複雜度。
二、預測結果：以支撐向量機系統預測效果最佳，其次是類神經網路，再者為傳統統計之Logit預測模式。綜合而言，人工智慧預測方法準確度優於傳統統計方法。

　　支撐向量機方法有完備的數學理論為基礎，能充份實現風險最小化原則，具備優良的學習推廣性能，運用於二分類樣本上，成效較其它兩種方法為佳。銀行授信管理階層可依企業相關財務指標及相關財務訊息推判公司經營績效與財務體質，減少超額授信與逾期放款的情形發生。因此，本研究期望藉由導入支撐向量機技術，增進人工智慧方法在銀行授信決策上之良好預測成效，供銀行業者做為授信業務之參考。

　　Under more and more competitive environment, the profits earned by financial institutions are getting thinner. Interests obtained by banks are gradually not enough to defray the loss on bad accounts. In the risk management field, a good early warning system with clearly loaning rules will help the bank’s accommodator to make loans.

　　Loan evaluation is an important issue for financial institution. With the well development of computer science and artificial intelligence, many of the operating obstacles can be overcomed. Based on theories of Support Vector Machines and Neural Networks, a model is set up to analyze and forecast the debit side’s credit and verify the accuracy of models. Empirical results are as follows:

1.In terms of the precision of prediction, Support Vector Machines leads to the highest accuracy, compared to Neural Networks which ranks the second, and Logit analysis is the lowest among the three.
2.With respect to variables choosing, Factor Analysis will effectively extract the important factors.

　　With complete mathematics theories, Support Vector Machines can comply with the principle of risk minimization and possess better performance. The results indicate that SVM is better than the others. It is anticipated that well applying SVM theory to loaning decision of the bank can be of much usefulness.

[1]陳肇榮(1983)，「運用財務比率預測財務危機之實證研究」，國立政治大學企業管理研究所博士論文。
[2]陳明賢(1986)，「財務危機預測之計量分析研究」，國立台灣大學商學研究所未出版碩士論文。
[3]黃小玉(1988)，「銀行放款信用評估模式之研究－最佳模式之選擇」，淡江大學管理科學研究所未出版碩士論文。
[4]周國政(1989)，「專家系統在財務診斷上應用之研究」，國立中山大學企業管理研究所未出版碩士論文。
[5]許正春(1989)，「銀行授信考量因素之研究」，國立交通大學管理研究所未出版碩士論文。
[6]陳文生(1989)，「財務分析應用於銀行放款信用評估之研究」，國立中山大學企業管理研究所碩士論文。
[7]邱志榮(1990)，「公司營運危機之預測－財務比率與現金流量比率之比較」，國立成功大學工業管理研究所未出版碩士論文。
[8]洪文慶(1994)，「銀行對借款者經營績效評鑑與信用評等之研究」，淡江大學管理科學研究所未出版碩士論文。
[9]蔡秋田(1994)，「運用類神經網路預測上市公司營運困難之研究」，國立成功大學會計學研究所未出版碩士論文。
[10]羅際棠(1996)，「銀行授信與經營」，三民書局出版。
[11]張瑜真(1996)，「類神經網路在銀行授信風險之應用」，國立交通大學管理科學研究所未出版碩士論文。
[12]馮志剛(1996)，「公營與民營銀行個人擔保與信用放款授信評估之研究」，高雄工學院管理科學系未出版碩士論文。
[13]邱泰為(1996)，「企業信用評等評估模式之研究」，國立成功大學工業管理研究所未出版碩士論文。
[14]楊政麟(1997)，「運用類神經網路於股價指數套利之研究」，國立台灣科技大學管理技術研究所未出版碩士論文。
[15]蔣松原(1997)，「應用模糊理論與類神經網路於銀行授信決策模式之研究－以台灣上市公司為例」，國立中興大學統計學研究所未出版碩士論文。
[16]陳錦村、許通安、林蔓蓁(1996)，「銀行授信客戶違約風險之預測」，管理科學學報，第十三卷第二期，p.173-195。
[17]葉英俊、紀榮年(1998)，「企業評等與銀行授信」，金融研訓中心出版。
[18]夏郭賢、吳漢雄(1998)，「灰關聯分析之線性數據前處理探討」，灰色系統學刊，第一卷，第一期，第47-53 頁。
[19]葉國興、黃天麟、倪成彬(1999)，「銀行對企業授信規範」，財團法人金融人員研究訓練中心。
[20]葉怡成(2000)，「類神經網路模式應用與實作」，儒林出版社。
[21]陳順宇(2000)，多變量分析二版，華泰書局。
[22]林建州(2001)，「銀行個人消費信用貸款授信風險評估模式之研究」，國立中山大學財務管理研究所未出版碩士論文。
[23]陳家彬、賴怡洵(2001)，「台灣地區銀行放款有無擔保之決定因素-Logit模型之實證分析」，管理評論，第二十期第一卷，p.129-159。
[24]盧靜怡(2001)，「企業經營績效排名之預測－灰色關聯分析與類神經網路之應用」，國立台灣科技大學資訊管理系未出版碩士論文。
[25]黃盟樺(2002)，「以支撐向量機建構企業與個人財務體質評估決策之研究」，國立高雄第一科技大學資訊管理所未出版碩士論文。
[26]張勝春(2002)，「類神經網路在銀行授信決策之應用」，朝陽科技大學財務金融系未出版碩士論文。
[27]周文賢(2002)，「多變量統計分析SAS/STAT使用方法」，智勝出版社。
[28]陳勇誠(2003)，「銀行授信評等模式－以小額信貸為例」，雲林科技大學財務金融系未出版碩士論文。
[29]卜怡君(2003)，「運用灰關聯聚類與適應性類神經模糊推論系統於本國銀行業之信用評等」，國立台灣科技大學資訊管理系未出版碩士論文。
[30]徐美珍(2003)，「企業財務危機之預測」，國立政治大學統計學系未出版碩士論文。
[31]李羿儒(2003)，「財務危機預警模式之再探討－運用支撐向量機及Logit模式」，國立高雄第一科技大學財務管理所未出版碩士論文。
[32]葉怡芳(2003)，「構建財務危機預警模式—支向機與羅吉斯之應用，元智大學會計學系未出版碩士論文。
[33]劉慶平(2003)，「神經網路與支持向量機學習算法之理論與仿真研究」，河北燕山大學碩士論文。
[34]劉艷芳(2003)，「基於神經網路和支持向量機的暫態穩定方法研究」，華北電力大學碩士論文。
[35]張斐章、張麗秋、黃浩倫 (2004)，「類神經網路理論與實務」，東華書局出版社。

英文部份
[1]Altman, E.I., (1968), “Financial Ratios, Discriminant Analysis and the Prediction of Corporate Bankruptcy,” Journal of Finance, pp.589-609.
[2]Beaver, W.H., (1966), “Financial Ratios as Predictors of Failure,” Journal of Accounting Research, Vol. 4, pp.71-102.
[3]Blum, M., (1974), “Failing in Company Discriminant Analysis,” Journal of Accounting Research, Vol. 12, pp.1-25.
[4]Cronbach, L. J. and Gleser, Goldine C., (1957), “Psychological tests and personnel decisions,” Urbana: Univer. Illinois Press
[5]Chester D., (1990), “Why Two Hidden Layer are Better then One,” IEEE International Joint Conference on Neural Networks. Washington, DC. pp.265-268.
[6]Chih-Chung Chang and Chih-Jen Lin, LIBSVM: a library for support vector machines, 2001. http://www.csie.ntu.edu.tw/~cjlin/libsvm
[7]Cristianini, N., and J. Shawe Tayor, (2000), “An Introduction to Support Vector Machines,” Cambridge University Press.
[8]Cuieford, J. P. (1965). “Fundamental statistics in psychology and education,” N.Y.: Mcgraw-Hill.
[9]G. Cybenko, (1989), “Approximation by superpositions of a sigmoidal function,” Mathematics of Control, Signals, and Systems, Vol. 2, pp.303-314.
[10]Dawson C. W. and Wilby R. L., (2001), “Hydrological Modeling Using Artificial Neural Networks,” Progress in Physical Geography, Vol. 25, pp.80-109.
[11]Fan, A., and M. Palaniswami, (2000), “Selecting Bankruptcy Predictors Using a Support Vector Machine Approach,” Proceeding of the IEEE-INNS-ENNS International Joint Conference on Neural Networks, 6, pp.354-359.
[12]Girosi F, Poggio T, Caprile B., (1991),” Extensions of A Theory of Networks for Approximations and Learning: Outliers and Negative Examples. In: Lippmann R P, Moody J E, Touretzky D S eds,” Advances in Neural Information Processing Systems, Vol. 3, pp.750~756.
[13]Gorr, W.L., (1994), “Neural Networks in Forecasting : Special Section,” International Journal of Forecasting, pp.1-4.
[14]Hair, J. F., Anderson, R. E., Tatham, R. L., and Black, W. C., (1998), “Multivariate data analysis,” Upper Saddle River, NJ: Prentice-Hall
[15]Hartman E J, Keeler J D, Kowalski J M., (1990), “Layered Neural Networks with Gaussian Hidden Units as Universal Approximations”. Neural Computation, Vol.2, No.2, pp. 210-215.
[16]Hayashi Y., Sakata M. and Gallant S. I., (1990), “Multi-Layer Versus Single-Layer Neural Networks and an Application to Reading Hand-Stamped Characters,” International Conference on Neural Networks. Paris. pp.781-784.
[17]Huang, (2003), “Credit rating analysis with support vector machines and neural networks: a market comparative study, Decision Support Systems, pp.543-558.
[18]Hush D. R. and Horne B. G., (1993), “Progress in Supervised Neural Networks: What’s New Since Lippmann,” IEEE Signal Processing Magazine, Vol. 10, pp.8-39.
[19]Johnsen, T. and Melicher, R., (1994), “Predicting Corporate Bankruptcy and Financial Distress: Information Value Added by Multinomial Logit Models”, Journal of Economics and Business, Vol. 46, pp.269-286.
[20]Kurkova V., (1992), “Kolmogorov’s Theorem and Multilayer Neural Networks,” Neural Networks, Vol. 5, pp.501-506.
[21]Lau, (1987), “A five -state financial distress prediction model,” Journal of Accounting Research,” Vol. 25, pp.127-138.
[22]Li, E. Y., (1994), “Artificial Neural Networks and Their Business Application,” Information & Management, Vol. 27, pp.303-313.
[23]Mensah, Y., (1984), “An Examination of the Stationarity of Multivariate Bankruptcy Prediction Models: A Methodological Study,” Journal of Accounting Research, Vol. 122, pp.380-395.
[24]Nunnaly, J., (1978), “Psychometric Theory,“ New York, NY: McGraw-Hill
[25]Odom, M.; Sharda, R., (1990), “A neural network model for bankruptcy prediction”, IEEE INNS IJCNN, Vol.2, pp.163-168.
[26]Ohlson, J. A., (1980), “Financial Ratios and The Probabilistic Prediction of Bankruptch,” Journal of Accounting Research, Vol. 18
[27]Orgler, YE, (1970), “A Credit Scoring Model for Commercial Loans,” Journal of. Money, Credit, and Banking, pp.435-445.
[28]Platt, H. D. and M. B. Platt(1990), “Development of A Class of Stable Predictive Variables: the Case Of Bankruptcy Prediction,” Journal of Business Finance and Accounting, pp.31-51.
[29]Rashmi Malhotra and D.K. Malhotra, (2003), “Evaluating Consumer Loans Using Neural Networks,” The Internation Journal of Management Science, pp.83-96.
[30]Straub, D. W., (1989), “Validating instruments in MIS reseach,” MIS Quarterly, Vol 13, pp.147-169.
[31]Tam, K. Y. and Kiang, M. Y., (1992), “Managerial Applications of Neural Networks: The Case of Bank Failure Predictions,” Management Science, Vol. 38, pp.926-947.
[32]Tsukuda, J. and Baba, S.I., (1994), “Prediction Japanese Corporate Bankruptcy In Term of Financial Data Using Neural Network,” Computer Industry Engineering, Vol. 27, pp.445-448.
[33]Van Gestel, T., B Baesens, J. Suykens, M. Espionza, D. E. Baestaens, Jan Vanthienen, and B. De Moor(2003), “Bankruptcy Prediction with Least Squares Support Vector Machine Classifiers,” IEEE International Conference on Computational Intelligence for Financial Engineering, pp.1-8.
[34]Vellido, A., P. J. G. Lisboa, et al. (1999). “Neural networks in business: a survey ofapplications (1992-1998),” Expert Systems with Applications, Vol. 51, pp.51-70.
[35]Wilson, R. L. and Sharda, R., (1994), “Bankruptcy Prediction Using Neural Networks,” Decision Support System, Vol. 11, pp.545-557.
[36]Zhang, G., Patuwo, B. E., and Hu, M. Y., (1998), “Forecasting with Artificial Neural Networks : The State of the Art,” International Journal of Forecasting, Vol. 14, pp35-62.
[37]Zmijewski, M. E., (1984), “Methodological Issues Related to the Estimation of financial Distress Prediction Models” Journal of Accounting Research, Vol.22, pp.59-82.