Article Details
  • Published Online:
    July  2025
  • Product Name:
    The IUP Journal of Accounting Research & Audit Practices
  • Product Type:
    Article
  • Product Code:
    IJARAP020725
  • DOI:
    10.71329/IUPJARAP/2025.24.3.44-56
  • Author Name:
    D Srinivasa Rao
  • Availability:
    YES
  • Subject/Domain:
    Finance
  • Download Format:
    PDF
  • Pages:
    44-56
Volume 24, Issue 3, July-September 2025
Corporate Bankruptcy Prediction: Evaluating the Effectiveness of Machine Learning Techniques for Addressing Class Imbalance
Abstract

The paper explores the challenges posed by class imbalance in corporate bankruptcy prediction and examines how machine learning (ML) techniques combined with data balancing techniques can address these issues. It explains how class imbalance affects the accuracy and reliability of predictive models and highlights the difficulty in identifying minority class instances (bankrupt companies) among a majority class of non-bankrupt entities. Using two popular bankruptcy datasets—Polish Companies Bankruptcy (2007- 2012) data and Taiwanese Bankruptcy Prediction (1999-2009) data—the effectiveness of two ML models, Balanced Random Forests (BRF) and Gradient Boosting with Class Weights to handle class imbalance problem, was evaluated. The results provide evidence of improvement in the performance accuracy of the classification model in the context of the two new techniques. The study contributes to the field of predictive modeling for corporate bankruptcy by offering actionable insights into strategies for addressing class imbalance. It aims to enhance the reliability and applicability of ML models in the context of financial risk assessment when faced with high skewness in data.

Introduction

Corporate bankruptcy refers to a legal process in which a business firm declares that it is unable to repay its debts. It has significant implications for various stakeholders, including creditors, shareholders, employees, and customers. It can lead to job losses, financial losses for investors, disruptions in supply chains, and changes in industry dynamics.