OMO BANK LOAN ELIGIBILITY CLASSIFICATION USING MACHINE LEARNING MODELS

Abstract

The banking sector is crucial to the global economy, with banks like Omo Bank playing a pivotal role in supporting economic growth by providing loans and facilitating financial transactions. However, the increasing demand for loans places pressure on banks to accurately evaluate loan eligibility, a process that is often time-consuming, error-prone, and inconsistent. This problem impacts the bank's risk assessment and loan decision-making. Although previous studies have attempted to classify loan eligibility using machine learning techniques, there has been limited effort to develop models tailored to the specific criteria of Omo Bank, and some studies have used small datasets. This study aims to develop a machine learning model to predict loan eligibility using a dataset from Omo Bank with approximately 1,999 records collected between 2010 and 2016 E.C. The methodology involves using tools such as StandardScaler to normalize data and dummy encoding to handle categorical variables. Employing ensemble machine learning algorithms, including Random Forest, XGBoost, Extra Tree Classifier, and Bagging Tree Classifier, the study classifies loan eligibility. The results demonstrate that the Random Forest model outperforms the other models, achieving an accuracy of 99.25%, thus indicating its suitability for loan eligibility classification within Omo Bank. This model can contribute to more efficient, consistent, and reliable credit decisions, benefiting the bank and its customers.