Using decision trees for data mining:a case study based on the computer resources survey report of the General of Budget, Accounting and Statistics, Executive Yuan, Taiwan

• Main Authors: Hsu,Sung-Chi, 許松棋
• Other Authors: Brandt Tso
• Format: Others
• Language: zh-TW
• Online Access: http://ndltd.ncl.edu.tw/handle/40931558265953387010

碩士 === 國防大學管理學院 === 資訊管理學系 === 98 === The electronic data process center of Directorate-General of Budge, Accounting and Statistics, Executive Yuan will investigate computer application in every walk of life at the end of each year. The investigate trend and attributing are a bit of different year by year. This research is with the investigation result data from 1999 to 2008, sieve the industry and service classification out and choosing different attributing, parameter to survey data rules with three decision tree tools. The data rules can be applied in relevant similar cases and get these results sorted out. Because the slightly insufficient sample data, the research get materials from other source, exceed 1000 samples, to avoid the result of study biased. Beside that the research use three kinds of decision tree to prospect and compare in order to examine the result of study. Decision support system can utilize the output of decision tree prospect out and attributing parameter input to yield judgment data or case for policymaker reference. An organization or the unit store relevant business accumulated information, experience and case in a system for a long time. Utilizing the management information system can preserve and manage a large number of information effectively. The purpose of the research is applying the prospect rule to Decision support system. Keywords: Decision tree、Data mining 、Data warehouse、Decision support system