Feature Identification for Credit Risk Based on Multi-Objective Evolutionary Clustering

doi:10.12005/orms.2022.0195

Abstract

Abstract: In credit risk identification, clustering algorithm is often used to identify the feature of risk, by distinguishing samples from different inner structure. However, credit risk data has high-dimensional characteristics. When dealing with credit risk data, there often exist flaws in conventional clustering methods, such as trapping in local optimum, being deteriorated by redundant features and poor robustness. Therefore, we look into the credit risk assessment problem, by establishing a tri-objective optimization model and then designing a decomposition-based multi-objective subspace clustering algorithm to solve it. A comprehensive experiment is also conducted to demonstrate the advantages of our proposed method. Then we summarize the key factors for the risk assessment.

Key words: credit risk, feature identification, multi-objective optimization, clustering algorithm

摘要： 在信用风险识别领域,聚类算法常被用于区分不同风险等级的样本并识别风险特征。然而该领域中通常面临高维数据处理问题,导致传统聚类算法存在不适应此类问题的缺陷:易陷入局部最优、受冗余特征干扰、鲁棒性不强等。采用高维信用风险数据,研究上市公司信用风险,建立信用风险特征识别的三目标优化模型,设计基于分解的多目标子空间聚类算法进行求解。通过算法的横向对比实验,展示了所提出的算法在聚类精度和鲁棒性方面的优势,并根据聚类算法的权重分配结果,归纳总结上市公司信用风险评估过程中应重点关注的指标。

关键词: 信用风险, 特征识别, 多目标优化, 聚类算法

CLC Number:

F832.4

LIU Chao, LI Yuan-rui, XIE Jing. Feature Identification for Credit Risk Based on Multi-Objective Evolutionary Clustering[J]. Operations Research and Management Science, 2022, 31(6): 147-153.

刘超, 李元睿, 谢菁. 基于多目标进化聚类的信用风险特征识别[J]. 运筹与管理, 2022, 31(6): 147-153.

References

[1] Cerchiello P, Giudici P, Nicola G. Twitter data models for bank risk contagion[J]. Neurocomputing, 2017, 264: 50-56.
[2] Li B, Chan K C C, Ou C, Sun R. Discovering public sentiment in social media for predicting stock movement of publicly listed companies[J]. Information Systems, 2017, 69: 81-92.
[3] 刘超,李元睿,姜超,等.中国证券公司系统性风险测度及演化特征研究——来自20家上市证券公司的数据[J].中国管理科学,2019(5):11-22.
[4] 吴成颂,汪翔宇.市场竞争、商业银行金融创新与银行业系统性风险——基于14家商业银行的实证研究[J].经济与管理评论,2019,35(2):120-129.
[5] Oreski S, Oreski G. Genetic algorithm-based heuristic for feature selection in credit risk assessment[J]. Expert systems with applications, 2014, 41(4): 2052-2064.
[6] De Andrés J, Lorca P, de Cos Juez F J, Sanchez-Lasheras F. Bankruptcy forecasting: a hybrid approach using fuzzy c-means clustering and multivariate adaptive regression splines(MARS)[J]. Expert Systems with Applications, 2011, 38(3): 1866-1875.
[7] 李元睿.多目标进化聚类算法及其在信用风险管理的应用[D].北京工业大学,2019.
[8] Rahmani M, Atia G K. Subspace clustering via optimal direction search[J]. IEEE Signal Processing Letters, 2017, 24(12): 1793-1797.
[9] Deng Z, Choi K S, Chung F L, Wang S. Enhanced soft subspace clustering integrating within-cluster and between-cluster information[J]. Pattern Recognition, 2010, 43(3): 767-781.
[10] Rahmani M, Atia G K. Innovation pursuit: a new approach to subspace clustering[J]. IEEE Transactions on Signal Processing, 2017, 65(23): 6276-6291.
[11] Javed S, Mahmood A, Bouwmans T, Jung S K. Background-foreground modeling based on spatiotemporal sparse subspace clustering[J]. IEEE Transactions on Image Processing, 2017, 26(12): 5840-5854.
[12] Groll L, Jakel J. A new convergence proof of fuzzy c-means[J]. IEEE Transactions on Fuzzy Systems, 2005, 13(5): 717-720.
[13] Maulik U, Bandyopadhyay S, Mukhopadhyay A. Multiobjective genetic algorithms for clustering: applications in data mining and bioinformatics[M]. Springer Berlin Heidelberg, 2011.
[14] Kriegel H P, Kröger P, Zimek A. Clustering high-dimensional data: a survey on subspace clustering, pattern-based clustering, and correlation clustering[J]. ACM Transactions on Knowledge Discovery from Data(TKDD), 2009, 3(1): 1-58.
[15] Masulli F, Rovetta S. Clustering high-dimensional data[C]//Revised Selected Papers of the First International Workshop on Clustering High——Dimensional Data. Springer-Verlag New York, Inc. 2012: 1-13.
[16] Jing L, Ng M K, Huang J Z. An entropy weighting k-means algorithm for subspace clustering of high-dimensional sparse data[J]. IEEE Transactions on knowledge and data engineering, 2007, 19(8): 1026-1041.
[17] Gordon A D, Vichi M. Fuzzy partition for fitting a set of partitions[J]. Psychometrika, 2001, 66(2): 229-247.
[18] Deb K, Jain H. An evolutionary many-objective optimization algorithm using reference-point-based nondominated sorting approach, part I: solving problems with box constraints[J]. IEEE Transactions on Evolutionary Computation, 2014, 18(4): 577-601.
[19] Jain H, Deb K. An evolutionary many-objective optimization algorithm Using reference-point based nondominated sorting approach, part II: handling constraints and extending to an adaptive approach[J]. IEEE Transactions on Evolutionary Computation, 2014, 18(4): 602-622.
[20] Mukhopadhyay A, Maulik U, Bandyopadhyay S, Coello C. Survey of multiobjective evolutionary algorithms for data mining: Part II[J]. IEEE Transactions on Evolutionary Computation, 2014, 18(1): 20-35.
[21] Mukhopadhyay A, Maulik U, Bandyopadhyay S. A survey of multiobjective evolutionary clustering[J]. ACM Computing Surveys (CSUR), 2015, 47(4): 61.
[22] Zhang Q, Li H. MOEA/D: A multiobjective evolutionary algorithm based on decomposition[M]. IEEE Press, 2007.
[23] 王娟.商业信用融资对股价崩盘风险的影响——基于债务治理效应视角[J].经济与管理评论,2019,35(3):110-121.
[24] Bezdek J C, Ehrlich R, Full W. FCM: The fuzzy c-means clustering algorithm[J]. Computers & Geosciences, 1984, 10(2-3): 191-203.
[25] Lloyd S. Least squares quantization in PCM[J]. IEEE transactions on information theory, 1982, 28(2): 129-137.
[26] Xia H, Zhuang J, Yu D. Novel soft subspace clustering with multi-objective evolutionary approach for high-dimensional data[J]. Pattern Recognition, 2013, 46(9): 2562-2575.
[27] Rand W M. Objective criteria for the evaluation of clustering methods[J]. Journal of the American Statistical association, 1971, 66(336): 846-850.
[28] Liu J, Mohammed J, Carter J, Ranka S. Distance-based clustering of CGH data[J]. Bioinformatics, 2006, 22(16): 1971-1978.
[29] Liu X, Cheng H M, Zhang Z Y. Evaluation of community detection methods[J]. IEEE Transactions on Knowledge and Data Engineering, 2019, 31(4): 1-11.