In this paper we propose an application of jumping emerging patterns (JEPs) to the classification of images. We define of a new type of patterns, namely the jumping emerging patterns with occurrence count (occJEPs), which allow reasoning in transaction databases with recurrent items. Such data is a frequently used representation of images, for which classification is one of the most important data mining problems that needs to be solved accurately and efficiently. We provide a formal definition of the new type of patterns, an outline of an algorithm for finding occJEPs and a comparison with other rule- and pattern-based classifiers for a selection of sample images. © 2008 Springer-Verlag Berlin Heidelberg.
CITATION STYLE
Kobyliński, Ł., & Walczak, K. (2008). Jumping emerging patterns with occurrence count in image classification. In Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) (Vol. 5012 LNAI, pp. 904–909). https://doi.org/10.1007/978-3-540-68125-0_91
Mendeley helps you to discover research relevant for your work.