Predicting Credit Risk of SMEs in Malaysia: Machine Learning vs Deep Learning

This study helps in predicting the credit risk of small and medium-sized (SMEs) by developing models using artificial intelligence algorithms: Machine Learning (ML) and Deep Learning (DL). This is demonstrated using four different prediction algorithms—Random Forest (RF), Neural Network (NN), Support Vector Machine (SVM) and Decision Tree (DT). The results show that DL model produced better prediction result: 86.5% classification accuracy, 86.1% of F1 score for Random Forest Model; 85.3% of classification accuracy, and 84.3% of F1 score for Neural Network Model. The result confirms that DL algorithm provides more accurate prediction leading to higher accuracy.