Fuzzy Clustering In The Bank Customer Segmentation

With the development of economy and society, market competition becomes more and more severe. In this environment, banking should pay more attention to considering the issue of how to win in this severe competition. As for the competition in banking, the focus is customer, especially those who bring great benefits to banking. As the 80-20 role said, the excellent customers who take the percentage of 20% of total amount provide 80% of the profits. Excellent customers are the source of the profits of banking. The bank which has large amount excellent customers will have a bright future of development. So banking should segment customers and take different plans to different customers. As the key point of popularize differential products, customer segmentation influence bank's development. Because there are much difference between different kinds of customers, the bank which only provides single product will attract less attention. As the most important resources, customers are the key point in the competition between different banks. So the bank which provides more attractive products will have more opportunity to win. As a result, banking should segment customers based on customers'differences.There are many approaches to segment customers. Different approaches have different use scopes. With the development of information technology, more and more banks use data mining technology to segment customers. Taking clustering as an example, traditional segment approaches have a few disadvantages. This thesis focuses on the application of fuzzy clustering approach in banking.This thesis analyzes the condition of customer relationship management and customer segmentation, compares common customer segmentation approaches, and summarizes their advantages and disadvantages. Then the thesis introduces the fuzzy clustering approach's advantage and how it can be used in customer segmentation. In chapter four, the thesis analyzes the key indexes of customer segmentation in banking, and then takes credit card as an example, chooses segmentation indexes and normalizes them. In chapter five, the thesis discusses how to design and realize a customer segmentation system based on fuzzy clustering, and takes real data to test the system. At last, the thesis makes a conclusion about fuzzy clustering approach's application in banking, and discusses the usage in the future.