Font Size: a A A

The Research On Commodity Association Rules Mining

Posted on:2016-07-01Degree:MasterType:Thesis
Country:ChinaCandidate:H ChenFull Text:PDF
GTID:2308330470476867Subject:Software engineering
Abstract/Summary:PDF Full Text Request
With the updating of science and technology, as a new direction of data mining,has been used in business, the rapid development of it offers more business opportunities, stimulate the rapid development of commercial economy. So choose goods Association rules research have more important significance. Into the existing association rules mining from large amounts of data to find the associative relationships between data, data mining area is the most extensive research topic. The Apriori algorithm can find the link between the data, by analyzing the relationship between the data, can predict the future of the relationship between the commodities of the trend.Based on the principle of data mining of association rules mining, try to use data mining techniques to predict goods related products between contact, make the data mining technology are increasingly common in our daily life.Data mining algorithms algorithm is to find the support of all the collections of the project, it may not be less than the algorithm of minimum support. The algorithm for the first time in the support of individual projects count and frequent items were identified. After each pass, seed setting frequent item sets found in the previous pass a new frequent item sets, potential and their actual support is to get the data of calculation, which means frequent item sets is certain, and they called the seeds for the next pass. Repeat this process until no new frequent item sets.According to the data mining can find the link between the data, by analyzing the relationship between the data, can predict the future development trend, so that practitioners to make the right decisions, improve business efficiency.
Keywords/Search Tags:Apriori algorithm, association rules, data mining, support degree, confidence degree
PDF Full Text Request
Related items