Optimization of Banking Loan Fraud Prediction using Supervised Learning Algorithm

सार

Every nation's banking sector serves as the foundation of its financial system and economy, influencing how that nation develops economically. A bank's performance in the current competitive environment is based on the new technology services it provides. The Indian banking sector has experienced tremendous growth since the implementation of the liberalization, globalization, and privatization policies. Using online banking services is one of the biggest financial innovations. This paper's loans are derived from data sets. Data collection: We use the Kaggle repository to gather data. Samples used for prediction is 614 samples and column names are loan ID, Gender, Married ,Dependent, Education, Self_Employed, Applicant Income, Coapplicant income, loan Amount etc. Lastly, the data set has been divided based on specific attributes. Finding variations in performance on particular segments of the data set is the aim of this kind of modification. The suggested method is applied to load data and is based on the gradient boosting (GB) technique. 93.13% training accuracy and 93.13% testing accuracy are offered by the suggested GB ML technique. The accuracy of the suggested method is 14.66% higher than that of RF and 11.10 percent higher than that of LR.