Dynamic ensemble classification for credit scoring using soft probability

被引:81
作者
Feng, Xiaodong [1 ]
Xiao, Zhi [1 ]
Zhong, Bo [2 ]
Qiu, Jing [1 ]
Dong, Yuanxiang [3 ]
机构
[1] Chongqing Univ, Sch Econ & Business Adm, Chongqing 400044, Peoples R China
[2] Chongqing Univ, Coll Math & Stat, Chongqing 400044, Peoples R China
[3] Shanxi Univ Finance & Econ, Sch Management Sci & Engn, Taiyuan 030006, Shanxi, Peoples R China
基金
中国国家自然科学基金;
关键词
Credit scoring; Dynamic ensemble classification; Selective ensemble; Soft probability; Machine learning; SUPPORT VECTOR MACHINE; NEURAL-NETWORK; BANKRUPTCY PREDICTION; CORPORATE BANKRUPTCY; RISK-ASSESSMENT; SELECTION; MODELS; CLASSIFIERS; PERFORMANCE; SYSTEMS;
D O I
10.1016/j.asoc.2018.01.021
中图分类号
TP18 [人工智能理论];
学科分类号
081104 ; 0812 ; 0835 ; 1405 ;
摘要
In recent years, classification ensembles or multiple classifier systems have been widely applied to credit scoring, and they achieve significantly better performance than individual classifiers do. Selective ensembles, an important part of this group of systems, are a promising field of research. However, none of them considers the relative costs of Type I error and Type II error for credit scoring when selecting classifiers, which bring higher risks for the financial institutions. Moreover, earlier dynamic selective ensembles usually select and combine classifiers for each test sample dynamically based on classifiers performance in the validation set, regardless of their behaviors in the testing set. To fill the gap and overcome the limitations, we propose a new dynamic ensemble classification method for credit scoring based on soft probability. In this method, the classifiers are first selected based on their classification ability and the relative costs of Type I error and Type II error in the validation set. With the selected classifiers, we combine different classifiers for the samples in the testing set based on their classification results to get an interval probability of default by using soft probability. The proposed method is compared with some well-known individual classifiers and ensemble classification methods, including five selective ensembles, for credit scoring by using ten real-world data sets and seven performance indicators. Through these analyses and statistical tests, the experimental results demonstrate the ability and efficiency of the proposed method to improve prediction performance against the benchmark models. (c) 2018 Elsevier B.V. All rights reserved.
引用
收藏
页码:139 / 151
页数:13
相关论文
共 62 条
[1]   An empirical comparison of conventional techniques, neural networks and the three stage hybrid Adaptive Neuro Fuzzy Inference System (ANFIS) model for credit scoring analysis: The case of Turkish credit card data [J].
Akkoc, Soner .
EUROPEAN JOURNAL OF OPERATIONAL RESEARCH, 2012, 222 (01) :168-178
[2]   Classifiers consensus system approach for credit scoring [J].
Ala'raj, Maher ;
Abbod, Maysam F. .
KNOWLEDGE-BASED SYSTEMS, 2016, 104 :89-105
[3]   FINANCIAL RATIOS, DISCRIMINANT ANALYSIS AND PREDICTION OF CORPORATE BANKRUPTCY [J].
ALTMAN, EI .
JOURNAL OF FINANCE, 1968, 23 (04) :589-609
[4]  
Baesens B, 2003, DEV INTELLIGENT SYST
[5]  
Banfield R. E., 2005, Information Fusion, V6, P49, DOI 10.1016/j.inffus.2004.04.005
[6]  
Basel III. BCBS, 2011, GLOB REG FRAM MOR RE
[7]   Bagging predictors [J].
Breiman, L .
MACHINE LEARNING, 1996, 24 (02) :123-140
[8]   Mining the customer credit using hybrid support vector machine technique [J].
Chen, Weimin ;
Ma, Chaoqun ;
Ma, Lin .
EXPERT SYSTEMS WITH APPLICATIONS, 2009, 36 (04) :7611-7616
[9]   The impact of preprocessing on data mining: An evaluation of classifier sensitivity in direct marketing [J].
Crone, Sven F. ;
Lessmann, Stefan ;
Stahlbock, Robert .
EUROPEAN JOURNAL OF OPERATIONAL RESEARCH, 2006, 173 (03) :781-800
[10]   Recent developments in consumer credit risk assessment [J].
Crook, Jonathan N. ;
Edelman, David B. ;
Thomas, Lyn C. .
EUROPEAN JOURNAL OF OPERATIONAL RESEARCH, 2007, 183 (03) :1447-1465