Abstract: Combining the features of the high tolerance of BP network to noisy data as well as the ability to classify the pattern not been trained,this paper discusses how to apply the BP network to realize classification in data mining,describes and analyses the corresponding BP algorithm,and carries out an illustrative analysis to the customer credit in the bank business.In this example,according to the loan,the numbers of loaning and the payment-in-time ratio,a model worked for classification of customer credit-rating is created.

Key words: neural network, back-propagation algorithm, data mining, classification