In recent years, the domestic emphasis on systemic risk prevention has continued to increase. Chinese credit department, which is a crucial department to avoid significant systemic risks, has a particular responsibility to manage and control personal credit risks related to the lives of the public. Therefore, strengthening the management of the individual credit department, especially to guard against personal credit risks, build a data-driven quantitative model for scoring applicants’ credit, and improve the personal credit identification and evaluation methods on this basis is a vital part of individual credit risk prevention and systemic financial risk. In this paper, we first study the index system of personal credit risk identification, standard personal credit risk identification models, and some data fusion techniques. Then, we build an individual credit risk identification model based on the data fusion method on the above theoretical basis. After that, the selection of indicators was carried out among 20 variables of German credit data. Finally, an empirical study was conducted to compare the effect of the two data fusion methods on the performance of the individual credit risk identification model. The defects of single models in the personal credit risk problem domain are apparent. However, data fusion technology can help classification models to gain better uniformity in accuracy and robustness, but it also needs to be improved in terms of interpretability and promotion.

Tian Wei, Shuai Li. Research on Personal Credit Risk Assessment Based on Data Fusion. Academic Journal of Business & Management (2023) Vol. 5, Issue 5: 62-74. https://doi.org/10.25236/AJBM.2023.050509.

[1] Wiginton J. C. (1980). A Note on the Comparison of Logit and Discriminant Models of Consumer Credit Behavior. Journal of Financial and Quantitative Analysis, 15(3), 757–770.

[2] Makowski M., & Sosnowski J. (1985). A Decision Support System for Planning and Controlling Agricultural Production with a Decentralized Management Structure. In M. Grauer, M. Thompson, & A. P. Wierzbicki (Eds.), Plural Rationality and Interactive Decision Processes (pp. 296–305). Springer.

[3] Quinlan J. R. (1986). Induction of decision trees. Machine Learning, 1(1), 81–106.

[4] Quinlan J. R. (1993). C4. 5: Programming for machine learning. Morgan Kauffmann, 38(48), 49.

[5] Su L., Li Q., Xu X., & Guo X. (2006). Data Fusion Algorithm for Sensor Network Based on D-S Evidence Theory. Journal of Chinese Computer Systems, 1321–1325.

[6] White F. E. (1991). Data fusion lexicon. Joint Directors of Labs Washington DC.

[7] Hall D. L., & Llinas J. (1997). An introduction to multisensor data fusion. Proceedings of the IEEE, 85(1), 6–23.

[8] Durrant-Whyte H. F. (1988). Sensor models and multisensor integration. The International Journal of Robotics Research, 7(6), 97–113.

[9] Dasarathy B. V. (1997). Sensor fusion potential exploitation-innovative architectures and illustrative applications. Proceedings of the IEEE, 85(1), 24–38.

[10] Luo R. C., Yih C.-C., & Su K. L. (2002). Multisensor fusion and integration: Approaches, applications, and future research directions. IEEE Sensors Journal, 2(2), 107–119.

[11] Castanedo F. (2013). A review of data fusion techniques. The Scientific World Journal, 2013.

[12] Breiman L. (1996). Bagging predictors. Machine Learning, 24, 123–140.

[13] REN Xiao, JIANG Minghui, CHE Kai, & WANG Shang. (2016). The research on methods of personal credit scoring combined model selection based on optimized index system. Journal of Harbin Institute of Technology, 67–71.