基于SVM的银行信贷风险评估模型研究
|
摘要
银行风险管理一直是国际国内金融界关注的焦点,入世后,国内银行面临着国内外银行的激烈竞争;国内银行由于信贷决策手段跟不上企业经营与市场信息的快速变化,导致不良资产率高、信贷资产安全性差、不良贷款比重大、贷款损失严重,隐含着相当大的危机。如何建立有效的企业信用评估体系和银行信贷风险评估模型,为银行决策提供科学的量化决策依据,全面降低不良贷款率,提高信贷资产质量,是国内银行共同面临的研究课题。因此,研究银行信贷风险评估具有重大现实意义。
目前,国内银行大多采用定性与定量相结合的风险量化模型对信贷风险进行评估,由于银行信贷风险等级的划分是一个多分类问题,这种建模技术直接影响信贷风险预测精确度和模型评估能力。本文在深入研究了国内外银行信贷风险评估采用的建模技术基础上,针对BP神经网络在建立信贷风险评估模型时存在的推广能力差和预测时间长等缺点,同时考虑到支持向量机可以处理一些多分类问题,采用支持向量机方法建立银行信贷风险评估模型。然而,支持向量机在处理多分类问题时多采用基于二值分类器的多分类器构造算法,这种算法在提取样本子集时仅考虑到二值分类而忽视与其他类之间联系,造成了类间信息丢失降低了分类精度,针对这一问题,本文提出了基于支持向量回归机的多分类器构造算法。该算法从决策函数入手把二值分类器扩展成三值分类器,增强了类间联系,解决了信息丢失问题;通过引入支持向量回归机代替三值分类器,解决了三值分类器存在的由于参数调整过多造成的计算代价大的问题。并采用中国银行滨河支行提供的数据,通过多组数据实验,将实际情况与该算法结果进行了对比验证。实验结果表明,该算法可以有效地提高预测精度,同时能解决计算代价大的问题,减少了占用的内存空间,提高了运行效率,缩短了训练时间,为银行信贷风险快速、有效的评估提供了更可靠的依据。
The bank risk management has been the focus which the international and domestic financial circles have paid attention to all the time. After joining the WTO, the domestic banks are faced with the severe competition from the domestic and international banks. Since the means of credit decision support in the domestic banks can’t keep up with the rapid changes in the business operation and market information, the proportion of the non-performing assets and loan is quite high, and the assets security of credit is poor, as well as the loan losses are serious. So a tremendous crisis is hided in the above-noted problems. A common research subject by domestic financial circles faced is how to establish an efficiency enterprise credit evaluation system and a good bank credit risk evaluation model, so as to provide the scientific quantized reference for making decisions in the banks, reduce non-performing loan rate in an all-round way and improve credit assets quality. Hence, it is great realistic significance to study the bank credit risk evaluation.
The risk quantization model combining qualitative analysis with quantitative analysis is adopted by most banks to evaluate the credit risk currently. The risk rank division of the bank credit is a problem of multi-classed. The accuracy of credit predicted and the ability of model evaluating are affected directly by the modeling techniques. This paper makes deeply research on the modeling techniques of bank credit risk evaluation adopted by international and domestic. Aiming at the poor Generalization Ability and the long prediction time in using Back-Propagation NN to establish the model of bank credit risk evaluation, in the meantime, considering that Support Vector Machine can solve some problems of multi-classed; the paper adopts SVM to establish this model. However, the algorithm about structuring multi-classifier based on bi-classifier is adopted widely in using SVM to solve these problems. But the relation between two classes is considered only by the above-noted algorithm rather than the relation among other classes, so losing information among classes is inevitable. The information loss results in the decreases of accuracy classed necessarily. To improve the performance, the new algorithm about structuring multi-classifier based on Support Vector Regression is put forward in this paper. Bi-classifier is expanded to tri-classifier bringing about strengthen relation among classes and reduce information loss in the view of decision function, then tri-classifier is replaced by SVR so as to solve the problem of high computational cost caused by adjusting overabundant parameters. Finally, the facts and proposed algorithm results are compared by much test data provided by Bank of China binhe sub-branch. The experimental results show that the algorithm can improve efficiently the forecast accuracy, solve the problem of high computational cost, decrease the memory occupied, and increase efficiency. Therefore, it can offer more dependable reference for efficient and effective evaluation in bank credit risk.
目前,国内银行大多采用定性与定量相结合的风险量化模型对信贷风险进行评估,由于银行信贷风险等级的划分是一个多分类问题,这种建模技术直接影响信贷风险预测精确度和模型评估能力。本文在深入研究了国内外银行信贷风险评估采用的建模技术基础上,针对BP神经网络在建立信贷风险评估模型时存在的推广能力差和预测时间长等缺点,同时考虑到支持向量机可以处理一些多分类问题,采用支持向量机方法建立银行信贷风险评估模型。然而,支持向量机在处理多分类问题时多采用基于二值分类器的多分类器构造算法,这种算法在提取样本子集时仅考虑到二值分类而忽视与其他类之间联系,造成了类间信息丢失降低了分类精度,针对这一问题,本文提出了基于支持向量回归机的多分类器构造算法。该算法从决策函数入手把二值分类器扩展成三值分类器,增强了类间联系,解决了信息丢失问题;通过引入支持向量回归机代替三值分类器,解决了三值分类器存在的由于参数调整过多造成的计算代价大的问题。并采用中国银行滨河支行提供的数据,通过多组数据实验,将实际情况与该算法结果进行了对比验证。实验结果表明,该算法可以有效地提高预测精度,同时能解决计算代价大的问题,减少了占用的内存空间,提高了运行效率,缩短了训练时间,为银行信贷风险快速、有效的评估提供了更可靠的依据。
The bank risk management has been the focus which the international and domestic financial circles have paid attention to all the time. After joining the WTO, the domestic banks are faced with the severe competition from the domestic and international banks. Since the means of credit decision support in the domestic banks can’t keep up with the rapid changes in the business operation and market information, the proportion of the non-performing assets and loan is quite high, and the assets security of credit is poor, as well as the loan losses are serious. So a tremendous crisis is hided in the above-noted problems. A common research subject by domestic financial circles faced is how to establish an efficiency enterprise credit evaluation system and a good bank credit risk evaluation model, so as to provide the scientific quantized reference for making decisions in the banks, reduce non-performing loan rate in an all-round way and improve credit assets quality. Hence, it is great realistic significance to study the bank credit risk evaluation.
The risk quantization model combining qualitative analysis with quantitative analysis is adopted by most banks to evaluate the credit risk currently. The risk rank division of the bank credit is a problem of multi-classed. The accuracy of credit predicted and the ability of model evaluating are affected directly by the modeling techniques. This paper makes deeply research on the modeling techniques of bank credit risk evaluation adopted by international and domestic. Aiming at the poor Generalization Ability and the long prediction time in using Back-Propagation NN to establish the model of bank credit risk evaluation, in the meantime, considering that Support Vector Machine can solve some problems of multi-classed; the paper adopts SVM to establish this model. However, the algorithm about structuring multi-classifier based on bi-classifier is adopted widely in using SVM to solve these problems. But the relation between two classes is considered only by the above-noted algorithm rather than the relation among other classes, so losing information among classes is inevitable. The information loss results in the decreases of accuracy classed necessarily. To improve the performance, the new algorithm about structuring multi-classifier based on Support Vector Regression is put forward in this paper. Bi-classifier is expanded to tri-classifier bringing about strengthen relation among classes and reduce information loss in the view of decision function, then tri-classifier is replaced by SVR so as to solve the problem of high computational cost caused by adjusting overabundant parameters. Finally, the facts and proposed algorithm results are compared by much test data provided by Bank of China binhe sub-branch. The experimental results show that the algorithm can improve efficiently the forecast accuracy, solve the problem of high computational cost, decrease the memory occupied, and increase efficiency. Therefore, it can offer more dependable reference for efficient and effective evaluation in bank credit risk.
引文
[1]邱仁.MM商业银行信贷风险管理存在的问题及对策研究[D].四川大学,2004.
[2]朱兴德,冯铁军.基于GA神经网络的个人信用评估[J].系统工程理论与实践,2003,(12):70-75.
[3]徐晋.基于神经网络专家系统的创业企业信用等级评估研究[J].南京理工大学学报,2004.12,28(6),684-688.
[4]吴冲,吕静杰等.基于模糊神经网络的商业银行信用风险评估模型研究[J].系统工程理论与实践,2004,11,11:1-8.
[5]刘学伟,贺昌政.基于贝叶斯正则化神经网络的上市公司信用评价研究[J].软科学,2005,19(5).
[6]石庆炎.一个基于神经网络—Logistic回归的混合两阶段个人信用评分模型研究[J].统计研究,2005,5:45-49.
[7]甄彤,范艳峰.基于支持向量机的企业信用风险评研究[J].微电子与计算机,2003,23,5:136-139.
[8]赵晓翠,王来生.基于主成分分析和支持向量机的商业银行信贷风险评估[J].理论新探,2006,7.
[9]吴冲,王敏等,基于支持向量机的商业银行信用风险评估模型研究[J].预测,2005,1:52-55.
[10] Odom M D,Sharda R.A Neural Network Model for Bank-ruptcy Prediction[A].Pcroceedings of the IEEE International Joint Conference on Neural Networks[C].1990,2:163-168.
[11] Altman E,Marco G,Varetrto F.Corporate Distress Diagnosis;Comparisons Using Linear Discriminant Analysis and Neural Networks[J].Banking and Finance. 1994,18:505-529.
[12] Tam K,Kiang M.Managerial Applications of the Neural Networks;The Case of Bank Failure Predictions[J].Management Science,1992,38:416-430.
[13] Coats P,Pant L.Recognizing Financial Distress Patterns Using a Neural Network Tool[J].Management,1993,22(3):142-155.
[14] Piramuthu S.Financial Credit Risk Evaluation With Neural and Neurofuzzy Systems[J].European Journal of Operational Research,1999,112(2):310-321.
[15] Graepel Tetal.Classification on proximity data with LP-machines[C].Ninth International Conferenceon Artifical NeuralNetworks IEEE,London:Conference Publications, 1999, 470:304-309.
[16] Zhang X.Using Class-Center Vectors to Build Support Vector Machines[M].In Proceedings of NN SP’99,1999.
[17]刘江华,程君实,陈佳品.支持向量机训练算法综述[J].信息与控制,2002,31(1):45-49.
[18] Keerthi SS,et al. A Fast Iterative Nearest Point Algorthm for Support Vector Machine Classifier Design[C].India: Indian Institute of Science Banga2 lore,1999.
[19] Schoblkopf Bet al.New support Vector Algorithms[J].Neural Computation,2000,12:1083-1211.
[20]Yang M H,A huja N.A Geometric Approach to Train Support Vector Machines[C].Hilton Head Island;In proceedings of CV PR 2000,2000:430-437.
[21]Suykens J A KJ Vandewalle.Least Squares Support Vector Machine Classifiers[C].Neural Processing Letters,1999,9(3):293-300.
[22]阎辉,张学工,李衍达.基于核函数的最大间隔聚类算法[J].清华大学学报,2002,42(1):36-38.
[23]李红莲,王春花,袁保宗.一种改进的支持向量机NN-支持向量机[J].计算机学报,2003,26(8):1016-1018.
[24]阎满富,田英杰.改进的支持向量回归机[J].系统工程,2004,22(10):9-12.
[25]汪西莉,焦李成.一种基于马氏距离的支持向量快速提取算法[J].西安电子科技大学学报,2004,31(4):639-643.
[26]孔锐,张国宣,施泽生,郭立.基于核的K-均值聚类[J].计算机工程,2004,30(11):12-14.
[27]腾月阳,唐焕文,张海霞.一种新的支持向量机增量学习算法[J].计算机工程与应用,2004,36:76-80.
[28]孙宗海,孙优贤.用于回归估计的最小二乘广义支持向量机[J].系统工程理论与实践,2004,(7):94-98.
[29]张国宣,孔锐,施泽生,郭立,刘士建,薛明东.基于核聚类方法的多层次支持向量机分类树[J].控制与决策,2004,19(11):1305-1308.
[30]张春华,田英杰,张跃峰.支持向量分类方法理论基础的改进[J].运筹学学报,2004,8(2):66-71.
[31]刘华富.一种快速支持向量分类算法[J].长沙大学学报,2004,18(4):40-42.
[32]陈振州,李磊,姚正安.基于支持向量机的特征加权KNN算法[J].中山大学学报(自然科学报),2005,44(1):17-21.
[33]张英,苏宏业,褚健.基于数据域描述的模糊支持向量回归[J].信息与控制,2005,34(1):1-6.
[34]朱熹钢,王士同,杨静宇.鲁棒γ-支持向量回归机中参数γ的选择研究[J].控制与决策,2004,19(12):1382-1386.
[35]杨志民,邓乃扬.模糊线性支持向量回归机[J].计算机工程与应用,2004,36:54-57.
[36]李红莲,王春花,袁保宗,朱占辉.针对大规模训练集的支持向量机的学习策略[J].计算机学报,2004,27(5):715-718.
[37]王晓丹,王积勤.支持向量机研究与应用[J].空军工程大学学报,2004,5(3):49-56.
[38]魏宏业,吕永波,何葳,张仲义.基于数据挖掘的企业合作伙伴的寻求方法研究[J].管理科学学报,2004,7(1):60-65.
[39]张浩,赵相东,陆剑峰,张辉.数据挖掘在石化企业中的应用[J].计算机工程与应用,2004,30:208-210.
[40]张辉,张浩,徐征,陆剑峰.基于支持向量机的供应链伙伴企业选择方法的研究[J].计算机集成制造系统,2004,10(7):796-711.
[41]张辉,张浩,陆剑峰.支持向量机在流程型企业决策支持系统中的应用[J].计算机工程与应用,2004,23:209-211.
[42]陆剑雪,侍洪波.基于支持向量机的化工过程故障诊断[J].华东理工大学学报,2004,30(3):315-317.
[43]王李东,李志宇,文劲宇.基于SVR算法的短期快速预测研究[J].继电器,2005,33(9):17-21.
[44] ing-Feng Pai,Chih-Sheng Lin.A hybrid ARIMA and Support Vector Machines Model in Stock Price Forecasting[J]. The International Journal of Management Sience,2004:497-505.
[45] Ping-Feng Pai,Wei-Chiang Hong.Support Vector Machines with Simulated Annealing Algorithms in Electricity Load Forecast [J].Energy Conversation & Management,2005:2669-2688.
[46] Vapnik V N.Estimation of Dependencies Based on Empirical Data [M]. Berlin:Springer-Verlag, 1982.
[47] Trevor Hastie,Robert Tibshirani,erome Friedman. The Elements of Statistical Learning-Data Mining,Inference,and Prediction [M].Beijing:Publishing House of Electronics Industry,2004.
[48]邓乃扬,田英杰.数据挖掘中的新方法—支持向量机[M].北京:科学出版社,2004.
[49]邓乃扬等,无约束最优化计算方法[M].北京:科学出版社,1982.
[50]刘志刚等,支持向量机在多类分类问题中的推广[J].计算机工程与应用,2004,(7):10-13.
[51]赵晶等,基于支持向量机的多类形状识别系统[J].合肥工业大学学报,2004,27(1):23-26.
[52]黄勇,郑春颖等,多类支持向量机算法综述[J].计算技术与自动化,2005,12,24(4):61-63.
[53] Angulo.Learning with Kernel Machines into a Multi-Class Environment.Doctoral thesis,Technical University of Catalonia. In Spanish, 2001.
[54] Shashua,A.and Levin,A.: Taxonomy of large margin principle algorithms for ordinal regression problems. Neural Information Processing Systems,16 (2002).
[55] Angulo,C.: Learning with Kernel Machines into a Multi-Class Environment. Doctoral thesis, Technical University of Catalonia.In Spanish,2001.
[56]刘庆平.神经网络与支持向量机学习算法的理论及仿真研究[D].燕山大学,2004.
[2]朱兴德,冯铁军.基于GA神经网络的个人信用评估[J].系统工程理论与实践,2003,(12):70-75.
[3]徐晋.基于神经网络专家系统的创业企业信用等级评估研究[J].南京理工大学学报,2004.12,28(6),684-688.
[4]吴冲,吕静杰等.基于模糊神经网络的商业银行信用风险评估模型研究[J].系统工程理论与实践,2004,11,11:1-8.
[5]刘学伟,贺昌政.基于贝叶斯正则化神经网络的上市公司信用评价研究[J].软科学,2005,19(5).
[6]石庆炎.一个基于神经网络—Logistic回归的混合两阶段个人信用评分模型研究[J].统计研究,2005,5:45-49.
[7]甄彤,范艳峰.基于支持向量机的企业信用风险评研究[J].微电子与计算机,2003,23,5:136-139.
[8]赵晓翠,王来生.基于主成分分析和支持向量机的商业银行信贷风险评估[J].理论新探,2006,7.
[9]吴冲,王敏等,基于支持向量机的商业银行信用风险评估模型研究[J].预测,2005,1:52-55.
[10] Odom M D,Sharda R.A Neural Network Model for Bank-ruptcy Prediction[A].Pcroceedings of the IEEE International Joint Conference on Neural Networks[C].1990,2:163-168.
[11] Altman E,Marco G,Varetrto F.Corporate Distress Diagnosis;Comparisons Using Linear Discriminant Analysis and Neural Networks[J].Banking and Finance. 1994,18:505-529.
[12] Tam K,Kiang M.Managerial Applications of the Neural Networks;The Case of Bank Failure Predictions[J].Management Science,1992,38:416-430.
[13] Coats P,Pant L.Recognizing Financial Distress Patterns Using a Neural Network Tool[J].Management,1993,22(3):142-155.
[14] Piramuthu S.Financial Credit Risk Evaluation With Neural and Neurofuzzy Systems[J].European Journal of Operational Research,1999,112(2):310-321.
[15] Graepel Tetal.Classification on proximity data with LP-machines[C].Ninth International Conferenceon Artifical NeuralNetworks IEEE,London:Conference Publications, 1999, 470:304-309.
[16] Zhang X.Using Class-Center Vectors to Build Support Vector Machines[M].In Proceedings of NN SP’99,1999.
[17]刘江华,程君实,陈佳品.支持向量机训练算法综述[J].信息与控制,2002,31(1):45-49.
[18] Keerthi SS,et al. A Fast Iterative Nearest Point Algorthm for Support Vector Machine Classifier Design[C].India: Indian Institute of Science Banga2 lore,1999.
[19] Schoblkopf Bet al.New support Vector Algorithms[J].Neural Computation,2000,12:1083-1211.
[20]Yang M H,A huja N.A Geometric Approach to Train Support Vector Machines[C].Hilton Head Island;In proceedings of CV PR 2000,2000:430-437.
[21]Suykens J A KJ Vandewalle.Least Squares Support Vector Machine Classifiers[C].Neural Processing Letters,1999,9(3):293-300.
[22]阎辉,张学工,李衍达.基于核函数的最大间隔聚类算法[J].清华大学学报,2002,42(1):36-38.
[23]李红莲,王春花,袁保宗.一种改进的支持向量机NN-支持向量机[J].计算机学报,2003,26(8):1016-1018.
[24]阎满富,田英杰.改进的支持向量回归机[J].系统工程,2004,22(10):9-12.
[25]汪西莉,焦李成.一种基于马氏距离的支持向量快速提取算法[J].西安电子科技大学学报,2004,31(4):639-643.
[26]孔锐,张国宣,施泽生,郭立.基于核的K-均值聚类[J].计算机工程,2004,30(11):12-14.
[27]腾月阳,唐焕文,张海霞.一种新的支持向量机增量学习算法[J].计算机工程与应用,2004,36:76-80.
[28]孙宗海,孙优贤.用于回归估计的最小二乘广义支持向量机[J].系统工程理论与实践,2004,(7):94-98.
[29]张国宣,孔锐,施泽生,郭立,刘士建,薛明东.基于核聚类方法的多层次支持向量机分类树[J].控制与决策,2004,19(11):1305-1308.
[30]张春华,田英杰,张跃峰.支持向量分类方法理论基础的改进[J].运筹学学报,2004,8(2):66-71.
[31]刘华富.一种快速支持向量分类算法[J].长沙大学学报,2004,18(4):40-42.
[32]陈振州,李磊,姚正安.基于支持向量机的特征加权KNN算法[J].中山大学学报(自然科学报),2005,44(1):17-21.
[33]张英,苏宏业,褚健.基于数据域描述的模糊支持向量回归[J].信息与控制,2005,34(1):1-6.
[34]朱熹钢,王士同,杨静宇.鲁棒γ-支持向量回归机中参数γ的选择研究[J].控制与决策,2004,19(12):1382-1386.
[35]杨志民,邓乃扬.模糊线性支持向量回归机[J].计算机工程与应用,2004,36:54-57.
[36]李红莲,王春花,袁保宗,朱占辉.针对大规模训练集的支持向量机的学习策略[J].计算机学报,2004,27(5):715-718.
[37]王晓丹,王积勤.支持向量机研究与应用[J].空军工程大学学报,2004,5(3):49-56.
[38]魏宏业,吕永波,何葳,张仲义.基于数据挖掘的企业合作伙伴的寻求方法研究[J].管理科学学报,2004,7(1):60-65.
[39]张浩,赵相东,陆剑峰,张辉.数据挖掘在石化企业中的应用[J].计算机工程与应用,2004,30:208-210.
[40]张辉,张浩,徐征,陆剑峰.基于支持向量机的供应链伙伴企业选择方法的研究[J].计算机集成制造系统,2004,10(7):796-711.
[41]张辉,张浩,陆剑峰.支持向量机在流程型企业决策支持系统中的应用[J].计算机工程与应用,2004,23:209-211.
[42]陆剑雪,侍洪波.基于支持向量机的化工过程故障诊断[J].华东理工大学学报,2004,30(3):315-317.
[43]王李东,李志宇,文劲宇.基于SVR算法的短期快速预测研究[J].继电器,2005,33(9):17-21.
[44] ing-Feng Pai,Chih-Sheng Lin.A hybrid ARIMA and Support Vector Machines Model in Stock Price Forecasting[J]. The International Journal of Management Sience,2004:497-505.
[45] Ping-Feng Pai,Wei-Chiang Hong.Support Vector Machines with Simulated Annealing Algorithms in Electricity Load Forecast [J].Energy Conversation & Management,2005:2669-2688.
[46] Vapnik V N.Estimation of Dependencies Based on Empirical Data [M]. Berlin:Springer-Verlag, 1982.
[47] Trevor Hastie,Robert Tibshirani,erome Friedman. The Elements of Statistical Learning-Data Mining,Inference,and Prediction [M].Beijing:Publishing House of Electronics Industry,2004.
[48]邓乃扬,田英杰.数据挖掘中的新方法—支持向量机[M].北京:科学出版社,2004.
[49]邓乃扬等,无约束最优化计算方法[M].北京:科学出版社,1982.
[50]刘志刚等,支持向量机在多类分类问题中的推广[J].计算机工程与应用,2004,(7):10-13.
[51]赵晶等,基于支持向量机的多类形状识别系统[J].合肥工业大学学报,2004,27(1):23-26.
[52]黄勇,郑春颖等,多类支持向量机算法综述[J].计算技术与自动化,2005,12,24(4):61-63.
[53] Angulo.Learning with Kernel Machines into a Multi-Class Environment.Doctoral thesis,Technical University of Catalonia. In Spanish, 2001.
[54] Shashua,A.and Levin,A.: Taxonomy of large margin principle algorithms for ordinal regression problems. Neural Information Processing Systems,16 (2002).
[55] Angulo,C.: Learning with Kernel Machines into a Multi-Class Environment. Doctoral thesis, Technical University of Catalonia.In Spanish,2001.
[56]刘庆平.神经网络与支持向量机学习算法的理论及仿真研究[D].燕山大学,2004.
© 2004-2018 中国地质图书馆版权所有 京ICP备05064691号 京公网安备11010802017129号 地址：北京市海淀区学院路29号 邮编：100083 电话：办公室：(+86 10)66554848；文献借阅、咨询服务、科技查新：66554700 |