Study on Predictive Model of Air Cargo Customer Defection Based on Decision Tree

Along with the continuous expansion of the air cargo market, more and more cargo airlines have been established. Facing the challenges from domestic and foreign competitors, airlines need to understand customer demands and analyze their values properly in order to compete successfully in the field. And data mining analysis techniques can help the company to hold the demands of customers and provide scientific and technical support for air cargo developing strategies and decision-making. The using of data mining could not always stand on the level of algorithms and models. Data mining should connect with the air cargo business conditions so as to dig out useful rules. This thesis classifies the air cargo customers by the methods of decision making tree, so as to forecast customer´s leaving and making decision.