Influence of Artificial Intelligence and Machine Learning Approach on Optimizing Bank Lending Decisions

Abstract

Much inefficiency in traditional bank lending processes characterizes the modern financial landscape, hence a main goal should be a critical analysis of these shortcomings. Previous studies have pointed out a number of drawbacks, including inefficiencies in terms of time, high expenses, and problems with flexibility, inconsistent results, poor data identification, and absence of real-time processing. These elements increase the risk of credit flow and lead to liquidity mismatches, issues with repayment capacity, and an overemphasis on asset seizure. Furthermore, because the data is vast and diverse, it is difficult to locate and resolve these discrepancies inside financial networks. In response; this study promotes the use of machine learning and artificial intelligence-based statistical techniques to address current limitations, providing improved speed and accuracy in bank lending choices. Using machine learning algorithms, the research employs a thorough methodology to statistically analyze the banking behaviors of its customers. A comprehensive analysis is conducted on banking parameters like credit score, debt-to-income ratio, credit duration, repayment history, bank account history, loan amount, and collateral utilizing information from trustable public domain banking datasets. The current study suggests the Hybrid Random Forest based Grey Wolf Algorithm (RF-GWA) for processing bank loan data sets most quickly and accurately in terms of predictions. When it comes to processing bulk baking data sets and making lending recommendations, RF-GWA performed better than the current algorithms Artificial Neural Network (ANN), Deep Neural Network (DNN), Convolution Neural Network (CNN), and non-hybrid Random Forest Method (RFM). It achieved close to 97% accuracy, over 94% precision, over 98% recall, and over 95% cross validation.