Using a Hybrid Approach Based on Artificial Neural Networks and Rough Set Theory for Modeling Customers Brand Loyalty in Mobile Telecommunicating Industry

Customers brand loyalty is one of the key indicators that reflect the market power of corporate brand. The aim of this study was the modeling of customers brand loyalty and investigating its effective factors in mobile telecommunication industry. For this purpose a hybrid method based on Artificial Neural Networks and Rough Set Theory was used. Based on the research framework, in the first stage, effective factors on customers’ brand loyalty were identified with use of an extensive literature review. Required data were gathered through a questionnaire survey among 913 of Irancell and Hamrahe-Aval subscribers in Yazd province. In the next step, sets of most effective attributes in modeling customers’ brand loyalty were identified with use of rough set theory. Then, based on artificial neural network approach, the best performing attribute set was selected and used for modeling customers’ brand loyalty. Finally, after presenting final network model together with synapsis weighs, input variables have been ranked with use of sensitivity analysis. Results of this study indicated that brand credibility, users’ imagery and pricing plans are the most effective factors on customers’ brand loyalty.