麻豆精品视频在线_韩国三级av_亚洲一区中文_涩涩的网址_91极品尤物_日本猛少妇色xxxxx_在线性视频_亚洲第一综合_大牛影视剧免费播放在线_被触手肉干高h潮文

Research on a Fast Algorithm for Mining Commodity Association Rules

• 1、Management School, Wuhan University of Science and Technology, Wuhan 430070

Abstract：The traditional association rule data mining algorithm mostly uses the Apriori algorithm, which is sensitive to the minimum support and minimum confidence parameters, and cannot generate structured association rules, which is less efficient. On the basis of Apriorialgorithm, a fast association rule mining (FARM) algorithm is proposed, which adopts the proposed guide function dominance and dynamic combination number generation technology to realize the selection, connection and decomposition of frequent item sets, and solves the feasibility and efficiency of association rule mining. Through the experimental analysis of the UCI real dataset, it is found that the number of association rules mined by the FARM algorithm is almost not restricted by the minimum support degree, which effectively overcomes the problem of parameter sensitivity. The algorithm can accurately control the number of association rules by adjusting the guidance function parameters. In addition, the time performance of the FARM algorithm and the classical Apriori algorithm under different minimum support degrees is evaluated through comparative experiments. Experimental results show that the FARM algorithm can not only complete the mining of effective association rules in a very short time, but also avoid the parameter-sensitive problem at the same time.

Keywords： commodity association analysis association rules boot function number of dynamic combinations Apriori