New approach clustering algorithm for customer segmentation in automobile retailer industry
Journal Title: International Research Journal of Applied and Basic Sciences - Year 2012, Vol 3, Issue 7
Abstract
This research presents a new method for automatic customer segmentation based on usage behavior, without using a human specialist. In addition, the research focuses on profiling customers and finding a relation between the profile and the segments. The customer segments were constructed by applying modified Gustafson-Kessel clustering algorithm. The customer’s profile was based on personal information of the customers. A novel Support Vector Machines was used to estimate the segment of a customer based on his profile.To demonstrate the efficiency of the proposed method, this work performs an empirical study of a Nissan automobile retailer to segment over 5000 customers. This led to solutions for the customer segmentation with six segments.
Authors and Affiliations
Seyed Reza Bahari Saravi| M.S. Student of Industrial engineering, Amirkabir University of Technology, Tehran, Iran,arezohamzei@yahoo.com
Finite Element Modeling of Fracture Failure in Composite Materials
A Finite Element (FE) model is developed to predict the crack propagation in mode I and mode II tests. The progressive mode I and mode II delamination failures in unidirectional and multidirectional composite laminates a...
Earthquake-Resistant Strategies in Traditional Architecture of Tabriz
Tabriz, as a city located within a strategic location, northwest of Iran, has confronted many earthquakes throughout its long history. The influences of earthquakes on the old city of Tabriz are distinguished from other...
The Effect of Emailing on Motivation of Adult English Learner of Iran
This paper originally was motivated by the claim that using emailing in language institutes has strong effect on motivation of adult learners. It investigated the motivation orientation of the Iranian English learners to...
Using a Mathematical Multi-facility location Model for the Market Competition
A class of competitive facility location and allocation (routing) models on the market in which two competing companies successively open their facilities is considered in this paper. Each client chooses an open facility...
Study of drought with SPI index (case study: Ghareh Chai and Karkheh basins)
To geo location and climate condition of Iran, in recent years, some hydro meteorological problems have emerged for different parts of Iran because of Low rainfall, it’s inappropriate distribution and drought. The goal o...