A Data-driven Case-based Reasoning in Bankruptcy Prediction

Abstract

There has been intensive research regarding artificial intelligence (AI) models for predicting bankruptcy in recent years. In this study, a data-driven case-based reasoning (CBR) method is proposed for bankruptcy prediction, testing its performance with financial data from German companies. A comparison of the predictive abilities of both the proposed method and other modern data-driven methods is presented. The results show that the proposed automatic design CBR technology has a good predictive performance, overcoming the design drawback of the CBR approach, which highly depends on prior expert knowledge about the corresponding field. In addition, the empirical studies show the explainability of CBR in the decision-making process of bankruptcy prediction.

Keywords: Decision support system, Bankruptcy prediction, Case-based reasoning, Particle swarm optimization

JEL Classification: M1, G33, G24