Research on Mining Multi-Weights Supports Association Rules with Frequent Pattern Growth Algorithm

• Main Authors: Jer-Guang Gu, 古哲光
• Other Authors: Cheng-Ming Yang
• Format: Others
• Language: zh-TW
• Published: 2004
• Online Access: http://ndltd.ncl.edu.tw/handle/47216864497199737395

碩士 === 立德管理學院 === 應用資訊研究所 === 92 === Recently years, it is important to mine item’s association rules from large database due to increasing considerable quantity of data constantly. In the past Algorithm of mining item’s association rules, most of those focus on times of trading count, but not profit. The main defect of that Algorithm is not effective to mining. Therefore this reason, this program has to provide a new method for improving the rate of time between mining item’s times and profit, is called MWFP-Growth (Multi- Weights Support Frequent Patterns Growth). This program alters FP-Growth to be applied to Weight Algorithm Multiple Support frequent patterns growth. It is effective to improve the defect of the rate of time of the association rules.