•  
  •  
 

Abstract

The primary objective of any classification model is to categorize observations into two or more distinct groups to predict outcomes—such as determining whether a company is likely to remain solvent or face bankruptcy (default vs. non-default). The study focuses on predicting the future performance and behavioral patterns of companies by utilizing both parametric and non-parametric statistical methods. Parametric methods rely on specific assumptions regarding the distribution of data to estimate parameters for problem-solving. In contrast, non-parametric approaches, such as tree classification (Decision Trees), offer flexibility by not assuming a specific data distribution. By comparing these methodologies, the research aims to enhance the accuracy of financial forecasting, providing stakeholders with reliable tools to anticipate and mitigate the risks of corporate bankruptcy through data-driven insights.

DOI

10.33095/jeas.v14i49.1373

Subject Area

Statistical

First Page

295

Last Page

315

Rights

http://creativecommons.org/licenses/by-nc-nd/4.0

Share

COinS