A Predictive Model of Behavior of Purchasers of Iranian Brands by Applying the Genetic Algorithm in Optimization of Decision Tree: Electric Appliance Industry

The purpose of this study is to provide a predictive model of the Behavior of Iranian Brands’ Purchasers, which was carried out in two co-ordinated steps. In the first step, a 26-item questionnaire has been employed, which was developed through literature review. The comments of 856 Mazandarani customers on the purchase of Iranian and foreign brands were collected based on actual purchase behavior. After pre-processing of data, in order to achieve the optimal solution range, the genetic algorithm was used and effective characteristics were selected. In the second step, the purchase behavior of buyers of Iranian and foreign brands was analyzed using a decision tree and the optimal models, and IF-THEN rules of purchasers of domestic and foreign brands were obtained. The results showed that the perception of external threats, national production support and income, are main indicators in separating purchasers of Iranian and domestic brands, and the higher is the perception of external threats and national production support in the customer, the higher the tendency to domestic brands; meantime, the groups of people with higher income have more aptness to buy foreign brands.