Using Alternative Covering Design Algorithm to Valuate the Reputation of Life Insurance Customer

With the huge increment of insurance data mining information,neural network technology can support the enterprise decision—making and reduce investment risk.However,the traditional neural network has some weakness,for example,the slow convergence of learning process,the poor network performance,the local minimum value maybe have, and so on.In this paper,we use the Alternative Covering Design Algorithm,which can not only avoid the weakness above, but also have good function of classification to valuate the reputation of life insurance customer and predict the trend of cheat behavior of customer.The result predicts that the precision of this method is very high,and the speed is quick. It obtains good effect.