Comparative study on Loan Approval Prediction Utilizing Machine Learning Approaches

Abstract

Due to the recent advancements in the banking sector and the growing popularity of borrowing money, a large number of people are submitting applications for bank loans. However, the growing incidence of loan defaults is a major problem for the banking sector in this vibrant economy. As a result, it is becoming increasingly challenging for banking authorities to accurately evaluate loan requests and manage the risk of loan default. In the banking sector, the two most important questions are (i) How hazardous is the borrower? and (ii) Should we provide the borrower money given the risk? In order to overcome the above-described problems, this study presents a pair of machine learning algorithms that assess particular features in order to decide whether or not to offer a loan. This will help banking authorities select the best individuals from a pool of applications. This study provides a thorough evaluation that contrasts the two algorithms: Decision Trees and Random Forests. Both approaches were applied to the same dataset, and the outcomes showed that the method using Random Forests outperformed the algorithm known as Decision Tree by a wide margin. In the end, it appears that a random forest approach performs better than the Decision tree. DT and RF, with a $\mathbf{9 0 \%}$ confidence level, Between the two most widely used machine learning algorithms and two statistically insignificant groups, the unbiased sample T-test result of $\mathbf{p}=0.33$ ($\mathbf{p}\gt0.05$) shows a statistically significant agreement. Random Forests seem to be able to forecast loan acceptance more accurately than Decision Trees.

Cite this work

No public copy is stored for this record, so only the metadata is cited.

Stored only when both papers are in the corpus — 16 citations exist upstream.

Papers that cite the same references — the relational fallback while embeddings are still backfilling.