Journal Article

·2022 OPEN ACCESS

Prediction of financial distress of companies with artificial neural networks and decision trees models

Nezir Aydın YTU , Nida Sahin , Muhammet Deveci YTU , Dragan Pamučar

Machine Learning with Applications

Abstract

Operational failures are closely related to many interest groups within and outside of the companies. Businesses may face financial failure as a result of the market conditions of the economy as much as the internal factors while maintaining their activities. Failure in managing the risks they face in these circumstances can lead to bankruptcy. For this reason, companies should be able to foresee their failures and consider correct measures by analyzing their current situation. With this motivation, to estimate and classify the financial failures of companies operating in different sectors a model is constructed using artificial neural networks (ANN) and decision trees (DTs). Unique models have been developed for each sector and it is aimed to compare the correct classification of non-bankrupt rates based on sector and to determine the most important variables affecting the financial failures on sectoral basis. In this context, for the companies listed in the BIST, 25 financial ratios and 2 non-financial variable were selected from 240 companies operating in manufacturing, service and trade sectors. In the model, near-zero error value has been targeted and non-bankrupt and bankrupt companies in the model have been classified correctly.

Keywords

Bankruptcy Context (archaeology) Business Artificial neural network Financial distress Variable (mathematics) Decision tree Finance Financial ratio Bankruptcy prediction Bank failure Value (mathematics) Actuarial science Computer science Artificial intelligence Financial system Machine learning Mathematics

Subject Areas

Financial Distress and Bankruptcy Prediction ·Accounting ·Social Sciences
Imbalanced Data Classification Techniques ·Artificial Intelligence ·Physical Sciences
Stock Market Forecasting Methods ·Management Science and Operations Research ·Social Sciences

Citations by Year