Combination Of Weak Learner And Strong On Stacking To Increase Bankruptcy Risk Prediction

Abstract

Bankruptcy is a fatal phenomenon for agencies. Minimizing the risk of bankruptcy can be done through prediction techniques using machine learning. Machine learning algorithms such as k-nearest neighbors can be used as modeling to generate accuracy values. Although in machine algorithms there are categories that distinguish between weak learners and strong learners. Both categories can be combined using the stacking ensemble method to improve accuracy performance. Such as using the k-nearest neighbors algorithm, decision tree, naïve Bayes, gradient boosting decision tree, light gradient boosting machine, and extreme gradient boosting machine. The results of research using this algorithm have an accuracy performance of 99.23%.