Prediction of Business Bankruptcy with the Help of Extreme Gradient Increase

Catalin-Emanuel Ciobota and Manuela-Violeta Tureatca
Additional contact information
Catalin-Emanuel Ciobota: Valahia University of Targoviste, Romania
Manuela-Violeta Tureatca: Dunarea de Jos University of Galati, Romania

Economics and Applied Informatics, 2022, issue 3, 178-185

Abstract: Financial institutions use business failure forecasting models to manage their investments. The accuracy of the forecast is an important factor in determining how much capital is needed to cover credit losses. The majority of studies use traditional statistical methods to model business failures (e.g., discriminant analysis and logistic regression). Models constructed are usually quite inaccurate, however. It is due to the imbalance between the classes of training samples (bankruptcies account for a small percentage of all firms) that are used to construct the models. Currently, various machine learning methods such as the random forest method and the gradient boosting method are widespread. In this study, the main focus is on using extreme gradient growth to predict bankruptcy. Extreme gradient boosting, using LASSO or Ridge regularization, penalizes complex models to avoid overfitting. Also, during training, extreme gradient boosting fills in the missing values in the data set depending on the amount of loss. In this article, in order to increase the efficiency of extreme degree growth, it is proposed to use SMOTE technology to improve class balance. The quality values of the solutions obtained by the improved extremal degree increase are compared with the solutions obtained by other methods.

Keywords: bankruptcy; enterprise; synthetic minority; insolvency (search for similar items in EconPapers)
Date: 2022
References: View references in EconPapers View complete reference list from CitEc
Citations:

Downloads: (external link)
http://eia.feaa.ugal.ro/images/eia/2022_3/CiobotaTureatca.pdf (application/pdf)

Related works:
This item may be available elsewhere in EconPapers: Search for items with the same title.

Export reference: BibTeX RIS (EndNote, ProCite, RefMan) HTML/Text

Persistent link: https://EconPapers.repec.org/RePEc:ddj:fseeai:y:2022:i:3:p:178-185

DOI: 10.35219/eai15840409301

Access Statistics for this article

More articles in Economics and Applied Informatics from "Dunarea de Jos" University of Galati, Faculty of Economics and Business Administration Contact information at EDIRC.
Bibliographic data for series maintained by Gianina Mihai ().