A Study on Corporate Credit Rating Model Using Ensemble Model Based on Error 


Vol. 47,  No. 1, pp. 98-112, Jan.  2022
10.7840/kics.2022.47.1.98


PDF
  Abstract

In order to study the improvement of the corporate credit rating model, the three single models of logistic regression, random forest, and gradient boosting and the error ensemble model combined logistic regression and gradient boosting were optimized and the discriminative power of each model was compared in this study. 1,295 companies out of 33,317 companies were used as samples. Each models are optimized 31 significant variables selected by reviewing 108 financial ratios. As a result of empirical analysis, error ensemble model had a limitation that it had excellent performance only in some indicators compared to random forest and gradient boosting. But error ensemble model"s overall performance was improved compared to logistic regression, and from the viewpoint of strengthening the interpretive power of logistic regression, its applicability as a credit rating model could be confirmed through this study.

  Statistics
Cumulative Counts from November, 2022
Multiple requests among the same browser session are counted as one view. If you mouse over a chart, the values of data points will be shown.


  Cite this article

[IEEE Style]

Y. Kim, D. Kim, J. Heo, G. Gim, "A Study on Corporate Credit Rating Model Using Ensemble Model Based on Error," The Journal of Korean Institute of Communications and Information Sciences, vol. 47, no. 1, pp. 98-112, 2022. DOI: 10.7840/kics.2022.47.1.98.

[ACM Style]

Yong-hwan Kim, Do-hyung Kim, Jae-hyuk Heo, and Gwang-yong Gim. 2022. A Study on Corporate Credit Rating Model Using Ensemble Model Based on Error. The Journal of Korean Institute of Communications and Information Sciences, 47, 1, (2022), 98-112. DOI: 10.7840/kics.2022.47.1.98.

[KICS Style]

Yong-hwan Kim, Do-hyung Kim, Jae-hyuk Heo, Gwang-yong Gim, "A Study on Corporate Credit Rating Model Using Ensemble Model Based on Error," The Journal of Korean Institute of Communications and Information Sciences, vol. 47, no. 1, pp. 98-112, 1. 2022. (https://doi.org/10.7840/kics.2022.47.1.98)

Search
(IN TITLE, AUTHOR, ABSTRACT,KEYWORDS)

Advanced Search ▼

POPULAR KEYWORDS
(TOP 10 KEYWORDS)

Recent Publications
(LAST 3 YEARS)






pISSN : 1226 - 4717
eISSN : 2287 - 3880