Analyzing Corporate Bankruptcy Prediction Models Using Machine Learning

Abstract

Corporate bankruptcy prediction is a critical and complex topic in finance and management that helps prevent financial crises and aids in better decision-making regarding financial risk management. In this regard, the use of machine learning algorithms, due to their ability to process complex data and provide accurate predictions, has emerged as a modern tool in this domain. This study aims to review and compare different machine learning models for corporate bankruptcy prediction and analyze the challenges and limitations of these models. This research follows a systematic literature review method, analyzing relevant scientific articles on the use of machine learning in bankruptcy prediction. Various models, including decision trees, random forests, artificial neural networks, and support vector machines, are examined, with their accuracy and performance evaluated based on empirical data. The results indicate that more complex models, such as neural networks and boosting algorithms, perform better than simpler models like decision trees when dealing with complex and multidimensional data. However, these advanced models face challenges such as interpretability issues, the need for large datasets, and complex parameter tuning. Machine learning techniques can provide high accuracy in predicting corporate bankruptcy, but improvements are needed in addressing challenges like data imbalance, enhancing data quality, and developing more interpretable models.