Decision Support System Terms of SME’s Credit Lending Based on Machine Learning Approach

Small and medium-sized (SMEs) companies have proven economic competencies and great potential in light of the recent global economic transformations in developed and developing countries, which qualified them to play a major role in development and reform policy, by providing job opportunities, intensifying the industrial fabric, and supporting large industries. It should be noted that these institutions need greater efforts and important components that respond to their aspirations and idiosyncrasies, to be able to survive and compete in the markets, as we have witnessed the bankruptcy and collapse of many companies due to financial conditions and obstacles. In this paper, we proposed building a lending decision support system (DSS) for SMEs and researching the reasons that prevent them from lending, and what are the pillars on which the credit decision in banks depends, and how difficult it is to analyze the variables and their information, the data set represents SMEs lending requests from Jordan Ahli Bank. Four machine learning algorithms were used to build an automated system to support decision-making in granting credit. The classification accuracy of the models was evaluated where the accuracy exceeded 90%, and SVM outperformed other rating models. Our findings show that the challenges that SMEs face to obtain credit are summarized by the lack of the company's capital, the recent date of incorporation, and the lack of guarantees.