篇名 | Classification Rules Mining of the Disabled Vehicle License Tax Exemption Cases in Taiwan |
---|---|
卷期 | 19:4 |
作者 | Yeh, Duen-yian 、 Cheng, Ching-hsue 、 Li, Kuo-yuan |
頁次 | 343-359 |
關鍵字 | Tax planning 、 vehicle tax exemption 、 disabled person 、 data mining 、 Scopus 、 TSSCI |
出刊日期 | 201412 |
DOI | 10.6126/APMR.2014.19.4.01 |
The purpose of this study was aimed at employing classification algorithms to mine specific rules from the disabled vehicle tax-exempt application cases. The inductive conclusions of classification results would be expected to provide the Taiwan government with a powerful assistance in overcoming the difficulties encountered. 17 important attribute parameters were selected and 19,002 valid samples were screened from Taiwan’s government database in 2010. Rough set theory, decision tree C4.5, Bayesian classifier and back propagation neural network were adopted with the aim at selecting a best classifier for this problem and obtaining the best classification results. The findings include: 1) Rough set theory was the best classifier with the classification accuracy, 94.39%; 2) 1,644 rules and nine key attribute parameters were discovered and used to generate rule conclusions for quickly and effectively checking application cases; and 3) Part of application cases were suggested to receive re-checking due to their doubtful situation.