<![CDATA[Abstract— Customers are important assets to support the passage of the Company. Company engaged in telecommunications have a high level of competition. To reduce the number of loyal customers who are not required analysis to predict the causes and out of its customers to be high. One way that can be used to analyze the database in large numbers to produce meaningful output is to use data mining. Prediction customer loyalty needed to predict the behavior of the customers most likely to cause customers to move to other competitors. To determine the predictive classification techniques used in data mining. In addition to predicting customer loyalty, in this paper will compare two data mining models are decision tree and Naive Bayes. Through testing dataset customers will get the best model to predict customer loyalty and gain a loyal customer forecasting. The measurement results show that the C4.5 algorithm is an algorithm that best predicted the customer loyalty that is equal to 88.14% by 0954 AUC values.]]>
Copyrights © 2014
