Emerging patterns: mining and applications
Kotagiri Ramamohanarao, James A Bailey · 2004
Emerging patterns are collections of items whose frequency changes significantly from one dataset to another. They are a means of discovering distinctions and contrasts between a collection datasets and have been shown to be a powerful method for constructing accurate classifiers. In this paper, we examine some different types of emerging patterns, look at techniques for mining them from large datasets and show they can be used for classification.