To Help Investors and Borrowers Make Informed Decisions in A Safer Environment A Study at UoB Uses AI To Predict Credit Scores in Lending

Sakhir – University of Bahrain (Khadijah Abdusalem)

26 October 2023

A student at the University of Bahrain (UoB) has developed a credit score prediction mechanism, using artificial intelligence (AI) algorithms, to help digital finance platform operators, investors, and borrowers make informed decisions, thus creating a safer and more effective environment for lending transactions (P2P).

The peer-to-peer (P2P) lending mechanism is known as a mechanism for lending money when the lender and borrower are equal parties (individuals and companies), without the need for the participation of banks or credit intermediaries, and the digital financing process occurs online through P2P platforms websites.

The study was prepared by the researcher in the Master’s program in Big Data Science and Analysis, student Fatima Mohammed Al-Muhaiza’a, and was titled: “A Comparative Study of Machine Learning Algorithms on The Data of Defaulters to Predict the Credit Score of P2P Lending Via Digital Financing Platforms.”.

Researcher Al-Muhaiza’a pointed out the study’s contribution to developing the credit score prediction mechanism in the P2P lending industry, by providing a systematic evaluation of machine learning algorithm models.

“After evaluating the predictive performance of machine learning algorithms, I concluded that the hybrid model consisting of a support vector machine (SVM), with the application of the data balancing technique (SMOTE), leads to the most accurate performance and outperforms the individual models, achieving 99%. Which is followed by the support vector machine (SVM) in terms of classification accuracy.”, Al-Muhaiza’a stated.

The supervisor of the thesis, Professor at the College of Information Technology, Dr. Nabil Mahmood Hewahi, pointed out that this study is an extension of previous studies, but what distinguishes it from others is that it used a hybrid system of SVM and SMOTE, which contributed to improving the accuracy of the results, as those using it will feel safe, and thus will have an increase in the financial flow, and an increase in investment and financial and commercial cooperation between the parties. This study also opens new horizons in scientific research to discover new methods, based on AI and machine learning, in the field of lending and credit scoring.

A specialized committee recently discussed the researcher’s thesis. The committee included the associate professor in the Department of Computer Science at UoB, Dr. Riadh Bin Mohammed Ksantini, as an internal examiner, and the professor at the College of Computer Science at the University of Windsor, Canada, Professor Dr. Boubakar Boufarma, as an external examiner, while the professor in the Department of Computer Science, College of Information Technology at UoB, Professor  Nabil Mahmood Hewahi, supervised the thesis

• 
• 
•