Predict the behavior of e-banking customers using data mining (Case study - Shahr Bank

Meisam Namazi¹

1) Shahid Beheshti University

IOI Code : XYAC-ZCAGC
Publication : 3rd.International Conference on Modern Research’s in Management,Economics & Accounting(3meaconf.com)
Abstract :
    Electronic banking is a subfolder of E-economy. The nature of intermediation in the banking sector is a multi-sector-based model. This extraction process model can analyze customer behavior and predict their future behavior. This study is based on the old model and extraction of e-banking customers at Shahr Bank. The proposed model in this study include the age, gender and education (Demographic characteristics) simplicity, versatility and security (website) and the types of banking transactions(Services). Period of study is 1390-1393. Data were analyzed using three methods: PNN, SOTA and Bayes. The results of all three methods are almost identical. Greatest customer Cluster is illiterate (elementary, illiterate and cycles) with almost .85% of transactions are defined for men and women is only 15% of transactions. It is expected that the illiterate cluster are applying for a loan in the future. The classification and clustering indicates that the nature of the mobilization and allocation of resources in the bank in the banking task, IT industry, electronic banking and virtual banking to the bank s ability to lower overhead costs to provide services to their customers.

Keywords : customer behavior sota pnn clustering e banking data mining