Microfinance Data Analysis in Banking Sector Using Data Mining Techniques

Abstract

Banks face lots of challenges associated with the bank loan, Nowadays there are many risks related to microfinance in bank sector. Every year, we face number of cases where people do not repay most of the microfinance amount to the banks which they cause huge losses. The risk associated with making decision on microfinance request approval is massive. In this study a classification model was built based on the microfinance data obtained from an agricultural bank of Sudan to predict the status of microfinance. The dataset has been preprocessed, reduced and made ready to provide efficient predictions. Random forest, NaiveBayes and KNN classification algorithms have been used to build the proposed model. By using Orange application the model has been implemented and tested. The accuracy for the above three techniques is Random forest 94.6%, NaiveBayes 87.4% and KNN 92.3%. Random forest selected as best algorithm based on accuracy. The final model is used for prediction with the test dataset and the experimental results proved the efficiency of the built model.