customer clustering based on rfm- cpi model using data mining techniques

Subject and purpose of the paper: the goal of this article is suggesting a new method for increasing the quality of customers clustering with increasing customer price index (cpi) to monetary variable (m). at this research, customers of one chain store of zahedan, iran according to rfm-cpi and rfm basic model variables and two-step algorithms are clustered to compare these two procedures. furthermore, determining the best method for customer clustering can be done. at this research, different steps of data mining and data analysis for discovering the knowledge of them were done according to the standard process of crisp-dm (1); this process includes system understanding, data understanding, data preparation, modeling, model assessment and deployment. according to the results, increasing the silhouette index at the rfm-cpi model in recent article in comparing with the basic rfm model defines high accuracy.this corrected model has advantages toward the main model; these advantages are contained: monetary changes at a period of time are identified, also according to clustering.