An Incremental Algorithm for Data Mining based on Rough Sets

Show simple item record Deng, Xiaofei 2011-04-21T21:49:09Z 2011-04-21T21:49:09Z 2011-04-02
dc.description.abstract In data mining, there are lots of methods about learning interesting rules (knowledge) from a database or an information table. Pawlak’s Rough Set Theory (RST), is a useful method for data mining and data analysis. Successful applications in data mining have proved that those learning approaches from the view of rough sets are rather helpful and valuable in obtaining interesting rules. Such approaches, however, assume that training examples recorded in a database will eventually converge to a stable state. That is, the information table is a finite and fixed set of records which share a common set of properties. In contrast to this assumption, the volume of data grows rapidly. For example, in 2006 the eBay’s massive oracle database has over 212 million registered users, holding two Petabytes of user Data. The database is running on Teradata with over 20 billion transactions per day. For management and market decision in such a business environment, an efficient rule learning algorithm with the real time processing ability is extraordinarily valuable. In this presentation, we will introduce an incremental RST algorithm based on the assumption that the objects in the information table change while time evolves. en_US
dc.language.iso en en_US
dc.publisher University of Regina Graduate Students' Association en_US
dc.relation.ispartofseries Session 5.5 en_US
dc.subject Data mining en_US
dc.subject Rough set theory en_US
dc.subject Incremental learning algorithm en_US
dc.title An Incremental Algorithm for Data Mining based on Rough Sets en_US
dc.type Presentation en_US
dc.description.authorstatus Student en_US
dc.description.peerreview yes en_US

Files in this item

This item appears in the following Collection(s)

Show simple item record

Search oURspace


My Account