Clustering Pattern of Life Insurance Customers (Case Study: An Insurance Company)

This research studies life insurance customers data in order to construct a clustering model for providing services. According to our estimations for sample size, our sample consists of 1000 life insurance policyholders who bought their policies in 2013 from an Insurance Company. Using clustering data mining models, effective factors and their relationships are examined. Finally, the results of different clustering models are compared to each other. According to the results, insurance companies can categorize life insurance customers into two main groups, including "profitable customers" and "risky customers". Hence, insurance company is able to offer suitable service packages to each group of customers. Some demographic indices, such as “gender” and “age”, and insurance indices, such as “annual premium” and “death rate due to accident” are considered as effective factors for identifying groups of customers.