Due to substantial technological advancements, peoples needs have expanded. Consequently, there has been an increase in the number of loan approval requests in the banking industry. Several criteria are considered while selecting a candidate for loan approval in order to ascertain the loans status. Banks encounter major challenges in evaluating loan applications and mitigating the risks linked to prospective borrower defaults. Due to the need to thoroughly assess the eligibility of every borrower for a loan, banks consider this process as notably burdensome. First the balancing of dataset will the performed. Recognizing the gravity of this task, the present study undertook the balancing of datasets as an imperative precursor, employing two oversampling techniques, SMOTE and ADASYN, for comparative analysis. The investigation aimed to discern the most efficacious balancing strategy for loan approvals by harnessing the analytical capabilities of algorithms such as Logistic Regression and Support Vector Machines (SVM). This exploration highlighted the distinct advantages and limitations intrinsic to each technique, underscoring the significance of aligning the choice with the datasets unique attributes and the financial institutions objectives. Compellingly, the results of the study demonstrated that SMOTE, when paired with SVM, emerged as the superior method, yielding the highest accuracy rate of 93.55%, thereby recommending its application as a robust and generalizable strategy for enhancing the accuracy and reliability of loan approval processes in the banking sector.