Font Size: a A A

The Research On Data Mining Applied To The Telecom

Posted on:2005-08-15Degree:MasterType:Thesis
Country:ChinaCandidate:M L ZhouFull Text:PDF
GTID:2168360122481252Subject:Computer application technology
Abstract/Summary:PDF Full Text Request
The fact that great deal of increased data are accumulated year by year with the computer more extensively applied makes the technology of data mining become more important. The technology of data mining, however, is also a new concept for every walk of life in China today. It can be forecasted that this technology will develop further in the not-so-distant future.First, commonly used arithmetic in data mining are listed in the paper, whose advantages and disadvantages and application at present are analyzed. Second, the detailed analyses of arithmetic that is related to my research into data mining are given, which involves the realization of arithmetic and the processing of experimental results. Thus, an improved decision-tree arithmetic is obtained and increasing by degrees study arithmetic is introduced, which can process increasing by degrees data and be used to the system design. Third, the importance of data mining applied to China's telecom enterprises is discussed. The following modes of three data mining systems, including client losing, client fraud, and time sequence, are developed to meet the need of Hangzhou telecom enterprise. Fourth, based on this, the point of view that data mining is applied to the telecom is put forward.The paper is divided into five chapters. In chapter one, the situation of data mining is summed up and the project of telcom is introduced. In chapter two, today's arilhmetic of data mining are introduced by the project of telcom. In chapter three, an emphasis is placed on the studying of the classified prediction arithmetic of data mining. And hence the CART arithmetic is mended comparing with the ID3 arithmetic. In chapter four, the survey of application of data mining to the telecom is conducted. In chapter five, the summarization of the paper is given.
Keywords/Search Tags:data mining, telecom, cart arithmetic, client losing, client fraud
PDF Full Text Request
Related items