Application of Machine Learning in Credit Scoring for Small and Medium Enterprises in Banking Sector

Thesis Abstract

Abstract
The banking sector continues to face challenges in effectively assessing credit risk for Small and Medium Enterprises (SMEs). Traditional credit scoring methods have limitations in accurately predicting the creditworthiness of SMEs due to their unique characteristics and limited financial history. This study investigates the application of machine learning techniques in credit scoring for SMEs to enhance the accuracy and efficiency of credit risk assessment in the banking sector. The research focuses on developing and evaluating machine learning models that leverage alternative data sources and advanced algorithms to improve credit scoring for SMEs. The study begins with a comprehensive review of the existing literature on credit scoring, machine learning, and SME financing to establish a theoretical foundation for the research. The literature review highlights the limitations of traditional credit scoring methods and the potential benefits of machine learning in enhancing credit risk assessment for SMEs. The research methodology section outlines the data collection process, model development, and evaluation criteria for the machine learning models. Through the analysis of a large dataset of SME financial and non-financial data, this study evaluates the performance of various machine learning algorithms, including logistic regression, random forest, support vector machines, and neural networks, in predicting creditworthiness for SME borrowers. The findings suggest that machine learning models outperform traditional credit scoring methods in terms of accuracy, sensitivity, and specificity for SME credit assessment. The discussion of the findings explores the factors influencing the predictive performance of machine learning models and identifies key variables that significantly impact credit scoring outcomes for SMEs. The study also examines the interpretability and explainability of machine learning models in credit scoring, addressing concerns related to model transparency and fairness in lending decisions. In conclusion, this research contributes to the existing literature by demonstrating the potential of machine learning in improving credit scoring for SMEs in the banking sector. The practical implications of adopting machine learning techniques for credit risk assessment are discussed, highlighting the benefits of enhanced risk management, reduced defaults, and increased access to finance for SMEs. The study concludes with recommendations for policymakers, financial institutions, and researchers to further explore the application of machine learning in credit scoring to support the growth and development of SMEs in the banking sector.