基于Metacost的客户信用评估半监督异构集成模型研究

doi:10.16381/j.cnki.issn1003-207x.2020.0933

中国管理科学 ›› 2022, Vol. 30 ›› Issue (12): 211-221.doi: 10.16381/j.cnki.issn1003-207x.2020.0933

• 论文 • 上一篇

基于Metacost的客户信用评估半监督异构集成模型研究

鄢澜¹, 李思涵², 肖毅³, 寇宇轩⁴, 刘敦虎⁵, 肖进²

1.四川大学出国留学人员培训部，四川成都610064； 2.四川大学商学院，四川成都610064；3.华中师范大学信息管理学院，湖北武汉430079； 4.澳门科技大学商学院，澳门特别行政区999078；5.成都信息工程大学管理学院, 四川成都610225

收稿日期:2020-05-20 修回日期:2020-07-03 发布日期:2023-01-10
通讯作者: 肖进(1983-)，男(汉族)，四川广安人，四川大学商学院，研究员，管理学博士，中国科学院数学与系统科学研究院博士后，研究方向：智能决策、商务智能、客户关系管理,Email：xjxiaojin@126.com. E-mail:xjxiaojin@126.com
基金资助:
国家自然科学基金资助面上项目(72171160，71974020)；四川省杰出青年基金资助项目(2020JDJQ0021)；四川省“天府万人计划项目”(0082204151153)；四川省软科学研究计划项目(2020JDR0120)；四川大学国家领军人才培育基金资助项目(sksyl2021-03)

Metacost Based Semi-supervised Heterogeneous Ensemble Model for Customer Credit Scoring

YAN Lan¹, LI Si-han², XIAO Yi³, KOU Yu-xuan⁴, LIU Dun-hu⁵, XIAO Jin²

1. Intensive Language Training Center, Sichuan University, Chengdu 610064, China;2. Business School, Sichuan University, Chengdu 610064, China;3. School of Information Management, Central China Normal University, Wuhan 430079, China;4. School of Business, Macau University of Science and Technology, Macau 999078, China;5. School of Management, Chengdu University of Information Technology, Chengdu 610225, China

Received:2020-05-20 Revised:2020-07-03 Published:2023-01-10
Contact: 肖进 E-mail:xjxiaojin@126.com

摘要/Abstract

摘要： 针对现实中信用评估存在的问题，本研究将元代价敏感学习、半监督学习和异构集成等技术结合，提出了基于Metacost的客户信用评估半监督异构集成模型(Metacost based semi-supervised heterogeneous ensemble model, Meta-Semi-HE)。该模型主要包括三个阶段：1)用Metacost方法修改初始有标签训练集得到Lm；2)在Lm上通过AdaBoost方法训练N个异构分类器hi(i = 1,…, N)，用伴随分类器组合Hi选择性标记无标签数据集的样本，并将其添加到Lm中，用新的Lm重新训练N个异构分类器。重复这一步骤，不断提高分类器性能，直至满足终止条件；3)用最终的N个异构分类器对测试集样本分类。在6个客户信用评估数据集上进行实证分析，结果表明，与已有的3种半监督集成模型和2种监督式集成模型相比，本研究提出的模型具有更好的客户信用评估性能。

关键词: 客户信用评估；类别分布不平衡；代价敏感学习；半监督；异构集成

Abstract: With the popularization of the credit business, effective risk aversion is one of the main means to maintain stable profits in the financial industry, and credit risk is one of the most common and important risk types in the financial industry. Therefore, accurate credit scoring of customers is very important. However, the class distribution of customer data used for credit-scoring models is often highly imbalanced, which means that there are significantly more customers with good credit as compared to customers with bad credit, and only a few customers who have successfully obtained loans can be labeled according to their future behavior, many customers who have applied for loans but failed to obtain them cannot be labeled. These characteristics bring great challenges to the establishment of scientific and accurate customer credit-scoring models, and existing researches cannot solve the above problems well. To make up for the lack of existing researches, meta cost-sensitive learning, semi-supervised learning, and heterogeneous ensemble learning are combined, and a Metacost based semi-supervised heterogeneous ensemble model (Meta-Semi-HE) is proposed for customer credit scoring. This model includes the following three stages: 1) Metacost is used to modify the initial labeled training set to obtain Lm; 2) N heterogeneous classifiers hi(i=1,…, N) are trained on Lm by AdaBoost, concomitant ensemble Hi is used to selectively mark samples of unlabeled data set, and adds them into Lm, N heterogeneous classifiers are retrained with the new Lm. Repeat this step to improve the performance of the member classifiers until the termination condition is satisfied; 3) the final trained classifiers are used to classify samples of the test set. The empirical analysis is conducted in six customer credit-scoring datasets, and the results show that the Meta-Semi-HE has better customer credit-scoring performance than the other five models in the evaluation criteria of AUC, f, Type I accuracy, and Type II accuracy. A new way of thinking for banks’ customer credit-scoring modeling is provided, which helps banks to avoid risks more effectively and promotes the healthy and stable development of credit business in the financial industry.

Key words: customer credit scoring; imbalanced class distribution; cost-sensitive learning; semi-supervised; heterogeneous ensemble

中图分类号:

TP18
F270

鄢澜, 李思涵, 肖毅, 寇宇轩, 刘敦虎, 肖进. 基于Metacost的客户信用评估半监督异构集成模型研究[J]. 中国管理科学, 2022, 30(12): 211-221.

YAN Lan, LI Si-han, XIAO Yi, KOU Yu-xuan, LIU Dun-hu, XIAO Jin. Metacost Based Semi-supervised Heterogeneous Ensemble Model for Customer Credit Scoring[J]. Chinese Journal of Management Science, 2022, 30(12): 211-221.

参考文献

［1］ Orgler Y E. A credit scoring model for commercial loans［J］. Journal of Money, Credit and Banking, 1970, 2(4): 435-445.
［2］ Zhou Hanhai, Wang Jinjin, Wu Jiadong, et al. Application of the hybrid SVM-KNN model for credit scoring［C］//Proceedings of the Ninth International Conference on Computational Intelligence and Security, Leshan, China, December 14-15, IEEE, 2013: 174-177.
［3］ Xia Yufei, Liu Chuanzhe, Li Yuying, et al. A boosted decision tree approach using Bayesian hyper-parameter optimization for credit scoring［J］. Expert Systems with Applications, 2017, 78(7): 225-241.
［4］ Wang Yadong, Jia Yanlin, Tian Yuhang, et al. Deep reinforcement learning with the confusion-matrix-based dynamic reward function for customer credit scoring［J］. Expert Systems with Applications, 2022, 200(14): 117013
［5］吴冲, 夏晗. 基于支持向量机集成的电子商务环境下客户信用评估模型研究［J］.中国管理科学, 2008, 16(S1): 362-367.Wu Chong, Xia Han. Study of customer credit evaluation under e-environment based on support vector machines ensemble［J］. Chinese Journal of Management Science, 2008, 16(S1): 362-367.
［6］王春峰, 康莉. 基于遗传规划方法的商业银行信用风险评估模型［J］. 系统工程理论与实践, 2001, 21(2): 73-79.Wang Chunfeng, Kang Li. A model based on genetic programming for credit risk assessment in commercial banks［J］. Systems Engineering Theory and Practice, 2001, 21(2): 73-79.
［7］ Xiao Jin, Wang Yadong, Chen Jing, et al. Impact of resampling methods and classification models on the imbalanced credit scoring problems［J］. Information Sciences, 2021, 569(8): 508-526.
［8］肖进, 刘潇潇, 谢玲, 等. 代价敏感的目标客户选择半监督集成模型研究［J］. 中国管理科学, 2018, 26(11): 186-196.Xiao Jin, Liu Xiaoxiao, Xie Ling, et al. A cost-sensitive semi-supervised ensemble model for customer targeting［J］. Chinese Journal of Management Science, 2018, 26(11): 186-196.
［9］ Domingos P. MetaCost: A general method for making classifiers cost-sensitive［C］//Proceedings of the Fifth ACM SIGKDD International Conference on Knowledge Discovery and Data Mining, San Diego, USA, August, ACM, 1999: 155-164.
［10］谢玲, 陈文婷，曹瀚文，等. 基于综合集成研讨厅的半监督客户关键特征选择模型研究［J］. 管理评论, 2020, 32(7): 236-245.Xie Ling, Chen Wenting, Cao Hanwen, et al. Semi-supervised key feature selection of customers based on hall for workshop of metasynthetic engineering［J］. Management Review, 2020, 32(7): 236-245.
［11］ Chawla N V, Lazarevic A, Hall L O, et al. SMOTEBoost: Improving prediction of the minority class in boosting［J］. Lecture Notes in Computer Science, 2003, 2838: 107-119.
［12］ Xiao Jin, Xie Ling, He Changzheng, et al. Dynamic classifier ensemble model for customer classification with imbalanced class distribution［J］. Expert Systems with Applications, 2012, 39(3): 3668-3675.
［13］ Xiao Jin, Zhou Xu, Zhong Yu, et al. Cost-sensitive semi-supervised selective ensemble model for customer credit scoring［J］. Knowledge-Based Systems, 2020, 189(2): 105-118.
［14］黄静, 薛书田, 肖进. 基于半监督学习的客户信用评估集成模型研究［J］. 软科学, 2017, 31(7): 131-134.Huang Jing, Xue Shutian, Xiao Jin. Semi-supervised learning based multiple classifiers ensemble model for customer credit scoring［J］. Soft Science, 2017, 31(7): 131-134.
［15］肖进, 薛书田, 黄静, 等. 客户信用评估半监督协同训练模型研究［J］. 中国管理科学, 2016, 24(6): 124-131.Xiao Jin, Xue Shutian, Huang Jing, et al. A semi-supervised cotraining model for customer credit scoring［J］. Chinese Journal of Management Science, 2016, 24(6): 124-131.
［16］肖进, 李思涵, 贺小舟, 等. 代价敏感的客户流失预测半监督集成模型研究［J］. 系统工程理论与实践, 2021, 41(1): 187-199.Xiao Jin, Li Sihan, He Xiaozhou, et al. Semi-supervised ensemble based on metacost model for customer churn prediction［J］. Systems Engineering Theory and Practice, 2021, 41(1): 187-199.
［17］ Blum A, Mitchell T. Combining labeled and unlabeled data with co-training［C］//Proceedings of the Eleventh Annual Conference on Computational Learning Theory, Madison, USA, July 24-26, Morgan Kaufmann, 1998: 92-100.
［18］ Blum A, Chawla S. Learning from labeled and unlabeled data using graph mincuts［C］//Proceedings of the Eighteenth International Conference on Machine Learning, Williamstown, USA, June 28-July 1, Morgan Kaufmann, 2001:19-26.
［19］张燕, 张晨光, 张夏欢. 基于改进图半监督学习的个人信用评估方法［J］. 计算机科学与探索, 2012, 6(5): 473-480.Zhang Yan, Zhang Chenguang, Zhang Xiahuan. Personal credit scoring method using improved graph based semi-supervised learning［J］. Journal of Frontiers of Computer Science and Technology, 2012, 6(5): 473-480.
［20］李凯, 崔丽娟. 集成学习算法的差异性及性能比较［J］. 计算机工程, 2008, 34(6): 35-37.Li Kai, Cui Lijuan. Diversity and performance comparison for ensemble learning algorithms［J］. Computer Engineering, 2008, 34(6): 35-37.
［21］ Bian Shun, Wang Wenjia. On diversity and accuracy of homogeneous and heterogeneous ensembles［J］. International Journal of Hybrid Intelligent Systems, 2007, 4(2): 103-128.
［22］韦艳艳, 李陶深, 刘美玲. 融合DECORATE的异构分类器集成算法［J］. 计算机应用研究, 2012, 29(11): 4134-4136.Wei Yanyan, Li Taoshen, Liu Meiling. Algorithm of heterogeneous classifiers ensembles based on DECORATE［J］. Application Research of Computers, 2012, 29(11): 4134-4136.
［23］ Zhou Zhihua, Li Ming. Tri-training: Exploiting unlabeled data using three classifiers［J］. IEEE Transactions on Knowledge and Data Engineering, 2005, 17(11):1529-1541.
［24］ Li Ming, Zhou Zhihua. Improve computer-aided diagnosis with machine learning techniques using undiagnosed samples［J］. IEEE Transactions on Systems, Man, and Cybernetics-Part A: Systems and Humans, 2007, 37(6): 1088-1098.
［25］ Li Yiyang, Su Lei, Chen Jun, et al. Semi-supervised learning for question classification in CAQ［J］. Natural Computing, 2017, 16(4): 567-577.
［26］ Tang Lingxiao, Fei Cai, Yao Ouyang. Applying a nonparametric random forest algorithm to assess the credit risk of the energy industry in China［J］. Technological Forecasting and Social Change, 2019, 144: 563-572.
［27］ Merz C J, Murphy P. UCI repository of machine learning databases［EB/OL］. 1995, http://www.ics.uci.edu/~mlearn /MLRepository.html.
［28］ Thomas L C, Edelman D B, Crook J N. Credit scoring and its applications［M］. Philadelphia: Society for Industrial and Applied Mathematics, 2002.
［29］ Linhart C, Harari G, Abramovich S, et al. PAKDD data mining competition 2009: New ways of using known methods［C］//Proceedings of the 13th Pacific-asia International Conference on Knowledge Discovery and Data Mining: New Frontiers in Applied Data Mining, Bangkok, Thailand, April 27-30, Springer, 2009: 99-105.
［30］ Liu Dongjiang, Liu Yanbi. An active learning algorithm for multi-class classification［J］. Pattern Analysis and Applications, 2019, 22(3): 1051-1063.

[1]	肖进, 刘潇潇, 谢玲, 刘敦虎, 黄静. 代价敏感的目标客户选择半监督集成模型研究[J]. 中国管理科学, 2018, 26(11): 186-196.
[2]	肖进, 薛书田, 黄静, 谢玲, 顾新. 客户信用评估半监督协同训练模型研究[J]. 中国管理科学, 2016, 24(6): 124-131.
[3]	肖进, 唐静, 刘敦虎, 谢玲, 汪寿阳. 基于改进GMDH的目标客户选择模型研究[J]. 中国管理科学, 2015, 23(10): 162-169.
[4]	朱帮助, 张秋菊, 邹昊飞, 魏一鸣. 基于OSA算法和GMDH网络集成的电子商务客户流失预测[J]. 中国管理科学, 2011, 19(5): 64-70.

基于Metacost的客户信用评估半监督异构集成模型研究

Metacost Based Semi-supervised Heterogeneous Ensemble Model for Customer Credit Scoring

PDF (PC)

可视化

摘要/Abstract

引用本文

使用本文

参考文献

相关文章 4

Metrics

本文评价

推荐阅读 0