A data prediction algorithm based on BP neural network in telecom industry

The data mining technology is more and more widely used in the telecom industry. But telecom data set always includes instances with missing values. Besides, many data mining models are sensitive for the missing value and distortion. Estimating missing values becomes an inherent problem. To address the problem, A prediction method is proposed for the missing value based on the BP neural network and K-means algorithm. The algorithm first clusters the dataset, and selects a certain cluster for the instance with missing values, then we train the BP net using the cluster and get the architecture parameters. When a test instance has a missing value of a certain attribute, we regard the attribute as the aiming attribute and apply the instance to the network, it will produce a prediction value . This paper uses real datasets from the telecom industry as the test datasets. The result shows that the algorithm can be used to predict the missing value of telecom industry with good performance.