Prediction of Bankruptcy of a company using machine learning techniques

Abstract

Bankruptcy prediction is an important modern technique because it helps not only the policy makers but the investors and all concerned people so they can take all necessary steps to prevent or to reduce the after effects of bankruptcy. Companies take a large quantity of loan and when they become bankrupt the entire system of the bank is shattered as it is really tough to get the given money bank. Hence, it is must needed to develop a system that could predict it the company’s bankruptcy and stop the loan. It is observed that the newer machine learning prediction model is outperforming the models of past thirteen years, using relatively small data but this data is still very big and balanced. Some of the machine learning models or techniques are Random Forest, XGBoost, Vector Machines. The main motive of the paper aims to compare the machine learning models which are support vector machine, random forest and XGBoost and give which one performs the best amongst them and give a novel accuracy.