International Journal of Computer Applications (0975 – 8887) Volume 82 – No 7, November 2013 13 A New Novel Cluster based Analysis of Bank's Customer Data with Self-Organized Map Networks Seyeyd Reza Khaze Department of Computer Engineering, Dehdasht Branch, Islamic Azad University, Dehdasht, Iran Emita Davoudi Takiyeh Department of Computer Engineering, Qazvin Branch, Islamic Azad University, Qazvin, Iran Isa Maleki Department of Computer Engineering, Dehdasht Branch, Islamic Azad University, Dehdasht, Iran Farhad Soleimanian Gharehchopogh Department of Computer Engineering, Hacettepe University, Ankara, Turkey ABSTRACT In the current global competing environment, creation of knowledge base and the use of it have been advantageous for the banks and the financial institutions and accounting and are being transformed to a strategic tool for competing among them and so data mining has been understood more and more in this field lately. In the today competing globe, banks and the financial institutions are trying to reach the advantage and be better than the others. Also, except execution of the business processes, the creation of the data knowledge and the use of it advantageous for the bank is being transformed to a strategic tool for competing. Taking into consideration this necessity, we have applied the Self-Organized Map (SOM) network in some cases of citizens in the banks of West Azerbaijan Province located at Republic Islamic of Iran. It is essential to cluster based solidarity analysis among of the specifications of customers to find common behavior points of them. However, it could be used to maintain the customers and find the new ones by the high responsible of programming of the banks. This approach leads to higher benefits and efficiencies in extracting and mining the likes and the wants of the customers. The results of the clustering analysis showed that the perspective of the customer about bank services and the effect of the electronic banking in banks selection, hold very similar junction patterns. Keywords Classification, Artificial Neural Network, Self-Organized Map, Learning, Algorithm. 1. INTRODUCTION The grouping algorithms and the data mining prediction algorithms have proved that can have better correction rate in comparison to the other statistical methods for predicting, grouping and clustering operations [1]. And taking into consideration that grouping and prediction are of the important procedures in the financial and accounting problems, data mining tools have proved themselves in this field very well. Of the fields the data mining tools are used, are the financial status prediction, cheating diagnosis management and credit risk estimation [2]. In the today competing globe, banks and the financial institutions are trying to achieve the advantage and be better than the others. Also, except execution of the business processes, the creation of the data knowledge and the use of it advantageous for the bank is being transformed to a strategic tool for competing. It must be taken into consideration that the ability of production and absorbing the data have increased considerably lately and must be taken into consideration that the information of these data are very important, so the accessibility of the data and the need for transforming them to the knowledge encourages the IT industry for using data mining. As the banking industry has faced many changes, now it has understood the need for data mining which helps it in competing market. The banks and the financial institutions have stepped on the way of data mining knowledge to find new customers and diagnose the cheating processes and service according to the shopping patterns of the customers [3]. The banks have been informed of the advantages of data mining in creation of the data mining for decision making and surely will benefit this and this will be advantageous in future competing. Lately the financial problems have been the challenge for the companies and predicting the break of the factories and are also a stressful point for the ones in relation with exchange market. As removing the regulations in the financial servicing industries and accepting the new wide technologies lead to the increase in competing in financial markets, and taking into consideration that the responsibility of the business strategies center of any company is to maintain the customers and also absorb the new ones, this knowledge could play important role in this struggle taking into consideration the abilities of data mining algorithms. In this paper, we have studied the solidarity among the specifications of the customers to find the common behavior points of them by the SOM networks, and the analysis can be used for maintaining the customers and finding the new ones by the high responsible people of the programming center of the banks. So, we have studied the abilities of the SOM networks in clustering. In the first section, the ex-activities are studied and then the clustering of the bank customers using the SOM networks is analyzed. 2. PREVIOUS WORK Data mining has played vast role in financial services, banks and business institutions lately which we study some here. Many of the data mining algorithms produce the lists of the regulations in which finding the beneficiary regulations depends on the user. Introducing the law selection mechanism, it is possible to use them for bank loans. Also production of law is possible by utilizing the decision tree C4.5 [4]. The results have showed that these methods of data mining are very effective in finding the favored laws. Also in Thailand, electronic banking is presented by different financial institutions which include business and governmental banks. On this perspective, electronic banking is very limited because of the limited use of the private data. The use of data mining techniques for analyzing the use of the