Designing a Data Mining Model Based on Predicting Factors Affecting the Improvement of Banking Operations (Case Study: Maskan Bank)

Abstract

Banks always evaluate various factors to improve their performance. Various research and studies have been carried out to improve the performance of banks. In this study, using data mining technique and combining two algorithms of random forest and Beding method, we have tried to evaluate the criteria for prediction of banks performance. Eight factors of employee productivity, deposit amounts and the number of deposit, the amount of loans and the number of loans facilities, satisfaction of electronic banking services, value added of housing price and percentage of loans for construction of housing have been used to predict the performance of Bank of Housing. The proposed method consists of randomly selected forest blending with Bagging method in 1, 5, 10 and 50 decision trees in each of three bags, suitable performance and appropriate and inappropriate performance, respectively. The results showed that the best result for 10 decision trees with 85.41% accuracy was obtained. The results were also evaluated by the results of other standard data mining algorithms. It was shown that the proposed method had better performance than other algorithms.