Multi-level Prediction of Financial Distress of Indian Companies Using Machine Learning

Title

Multi-level Prediction of Financial Distress of Indian Companies Using Machine Learning

Creator

Fernandes R.L.; Tantia V.

Description

Predicting Financial Distress (FD) and shielding companies from reaching that stage is vital, even indispensable for every business. FD, if not attended to on time, ultimately leads to bankruptcy. Prediction variables are essential to forecast the wreckage in the business; however, the prediction is successful when suitable models are used. This study aims to predict FD at three levels: from mild to severe, by applying a machine learning algorithm. The study identifies modern models using the machine learning approach for predicting multi-level FD and summarises the significance of modern models through machine learning technology, to sustain the future development of the economy. The modern models are free from rigid assumptions and have proved to be the best in the prediction of FD. The results show that FD prediction is important at multiple stages. The models performance will be high when the best features are selected using the Pearson Correlation and SFS Feature selection approach. Among the ten models used in the study, LightGBM Classifier shows the highest performance of 80.43% accuracy without feature selection. However, with Pearson Correlation Approach and SFS Feature Selection methods, the accuracy is 82.68% and 86.95% respectively. This study has major implications for the stakeholders of the company to take timely decisions on their investment and for the management as a yardstick to check the performance of the business. The Author(s), under exclusive license to Springer Nature Switzerland AG 2025.

Source

Studies in Systems, Decision and Control, Vol-555, pp. 265-274.

Date

2025-01-01

Publisher

Springer Science and Business Media Deutschland GmbH

Subject

Financial distress; Machine learning; Modern models; Multi-level; Performance measures; Prediction

Coverage

Fernandes R.L., Christ University, Hosur Road, Bangalore, 560029, India; Tantia V., Christ University, Hosur Road, Bangalore, 560029, India

Rights

Restricted Access

Relation

ISSN: 21984182

Format

Online

Language

English

Type

Book chapter

Identifier

https://doi.org/10.1007/978-3-031-67890-5_25

https://www.scopus.com/inward/record.uri?eid=2-s2.0-85210150952&doi=10.1007%2f978-3-031-67890-5_25&partnerID=40&md5=eda7d05f977bb495219f6150c474fa16