Customer Churned Analysis Based On Decision Trees Algorithm

Data mining is to discovery the interesting information and knowledge, which is useful, connotative, and unpredictable, from numerous data. Data mining is a technique that aims to analyze and understand large source data and reveal knowledge hidden in the data. It has been viewed as an important evolution in information processing. Why there have been more attentions to it from researchers or businessmen is due to the wide availability of huge amounts of data and imminent needs for turning such data into valuable information. During the past decade or over, the concepts and techniques on data mining have been presented, and some of them have been discussed in higher levels for the last few years.Customer Relationship Management (CRM), as a new application of data mining technologies in DSS recently, has made a breakthrough to the framework of traditional business management and merchandise planning. The traditional business strategy "all for production" is replaced by the brand-new concept "all for customer". Customer churned analysis is an important impact of CRM, which main technology is classification technique of data-mining. By the analysis of customer history data, we can draw the customer churned prediction model, which can be used to identify the latent churned customer and guide the customer reserve.C4.5 algorithm is one of the classification algorithms, which has a high precision and strong adaptability. Customer churned data has its own character: it has many continuous attributes and unbalanced distributing. Based on the proof of Fayyad and Irani, this paper advanced an improvement approach about the discretization and the penalty term of continuous-valued attributes. By a dynamic Under-sampling way, we can improve the unbalanced distributing. Base on the two improvements, this paper bring forward an improved algorithm icw_C4.5 of C4.5. Using the improved algorithm, the prediction accurate of object class is improved in a certain degree. At last, this paper uses the improved algorithm to build a model for a telecom churned dataset of a telecom company. The contrast analysis also proved the validity of the improved algorithm.