Research On Telecommunication Customer Credit Classification Based On Data Mining

In today's world,telecommunications operators in order to keep success in an increasingly competitive environment,not only to provide quality customer service and it is important to understand that customer's structure is necessary. In operations,customer credit rating can reflect the customer's basic structure information and help operators in the specific management practice to take positive and effective measures to reduce its credit risk and operational costs and improve its economic benefit.It is the most important work that correct and effective assessment of the credit rating in telecommunications customers. In this paper,we will use data mining techniques and SAS statistical analysis software to deal with the data of post-paid mobile phone users which from one of Chongqing telecommunications operators. predictions. At this stage,it is the major method to set up the credit range using the data of the consumer's monthly amount. This method is not very accurate in order to improve the credit rating of the forecast accuracy,this paper provide a new method of a credit rating trough the description of six parts of this article can be clear.First of all,this article describes the background and significance of the study; then the second part of thesis credit score ratings and the basic knowledge of data mining are summarized; And,the general idea of the study in the third part is clear the raw data to explore,and then use SOM clustering algorithm to customers to re-score rating to determine the type of label,classification and assess the new model using the C4.5 decision tree algorithm and testing its forecast effective,and compare to the original model in order to determine useful model in practice which carry out customer credit classification and prediction,after the latter part of the research paper to lay the foundation; Furthermore,papers in the fourth part analysis detailed the process of data pre-processing . It is the expansion of the pre-processing data stage; Part five is the core of this thesis,according to the third part of the thesis research ideas,based on the telecom operators to carry out the cases,and we usen the SAS data mining module analysis of the whole process and the evaluation indicators,the model results by the image of an intuitive graphical way to show out,while the model results in the management of applications,from credit management and marketing management for a two-pronged narrative; Finally,the sixth part of papers introduce the major results and shortcomings in order to help telecommunications operators in its future work to change the thinking style to bring a huge sense.