Prediction of customer churn using dynamic criterion based on Least Squares Support Vector Machine and Optimization Algorithm Ant Lion

In time of Global competition, life of the organization is depend on customer care. Avoid of leaving customer, in factof customer care. Understanding of customer behavior and on time detecting customer churn cause. Avoiding of losing customer. If we take preventive measures. Furthermore significantly reduces organization cost. It is because of for more expensive cost of attracting a new customer. Our goal in this paper is to predict the future behavior of our customers in two active or prone to churn categories by using the data in the database of the telecommunication industry of Iran. Studies carried out in last decade show that evaluation parameters usually are static and repetitive with various data mining techniques. In this research, using collected data and taking into account effective qualitative factors based on the quantitative parameters generated by the algorithm of the Least squares Support Vector Machine and Optimization Algorithm AntLion, tried to predict the loss of customers and calculate the efficiency of the suggested algorithm, compared to other algorithm proposed.

Keywords:

Customer churn prediction , customer relationship management , Least squares Support Vector Machine , Optimization Algorithm AntLion