Data mining extension for economics
Abstract
[ACCESS RESTRICTED TO THE UNIVERSITY OF MISSOURI AT AUTHOR'S REQUEST.] This research extends the fuzzy association rule mining technique capable of processing quantitative data and implements it to study the relationships among state social, political context, entrepreneurial activities, and economic development. The 5-cluster method is compared with 3-cluster method to study the effect of the number of clusters to this problem. Also, this research conducts a sensitivity analysis to study the effect of minimum support and confidence. The results of the analysis reveal that the 5-cluster method is necessary and the minimum support "0.05" and the minimum confidence "0.8" are chosen for the problem. Additionally, the marginal effect of the premises of a rule is generated automatically along with a set of interesting rules revealing the cause and effect of entrepreneurial activities. In economics, the improved fuzzy association rule mining methodology can be used as a complement of conventional econometric tools to help discovering previously hidden but potentially interesting economic relationships and generating ideas for new theories. In addition, it can be applied to various research and business fields to conduct quantitative data analysis.
Degree
M.S.
Thesis Department
Rights
Access is limited to the campus of the University of Missouri--Columbia.