Predicting Company Bankruptcy Using Random Forest Method

Abstract

This study aims to show the impact of machine learning as a field of artificial intelligence based on the fact computers can learn from data. Within this aim, our goal is to demonstrate how Random Forest outperforms other algorithms in our study. Predicting bankruptcy is becoming a necessary topic in finance and accounting activity. It has the potential to not only assist in making the best decision but also to reduce losses. Our dataset includes financial data from 6819 companies from 1999 to 2009. As a result, the Random Forest algorithm has given us 97.8% of accuracy in predicting whether a company will face bankruptcy or not.