COMPARATIVE STUDY OF SUPERVISED LEARNING IN CUSTOMER RELATIONSHIP MANAGEMENT

Customers are valuable to an organization. The competitive market environment makes customer relationship management very noteworthy for the business prospects. Therefore, research on customer relationship management is attracting data mining researchers. Data mining can support customer expansion by matching products with customers and better pursuing of product promotion campaigns. In this study, classification algorithms, namely J48, SGD, Bayes Net and Naïve Bayes Updatable were experimented on customer data. The comparison of these classification algorithms based on different performance metrics is presented. It will help to select a best suitable algorithm. The performance of the classification models is measured using 10-fold cross validation. The WEKA environment was utilized for the experiments and assessments of these methods. The 80% data were correctly classified by all these methods. It reveals that data mining, classification methods can be adopted for the customer relationship management study.