文档介绍：Identificatin of Salient Patterns for
Classification of Gene Expression Data
Gouchol Pok Guangri Quan Keun Ho Ryu
Yanbian University of Harbin Institute of Technology Chungbuk National University
Science and Technology School of Software DB/Bioinformatics Lab
Department puter Science Weihai, China Cheongju, Korea
Yanji, China ******@hit. ******@
gcpok2000@
Abstract—Identification of salient patterns for the classification disease-specific patterns called emerging patterns in the
of gene expression profiles is a useful step in examining the solid numeric values. Emerging patterns are discovered
biological significance and correlation of genes with disease using the border-based algorithm over the discretized values.
states. We propose a clustering-based approach in which In our approach, we exploit the advantageous points of
feature selection is first carried out to identify influential clustering methods which are useful to identify salient
genes and then salient patterns are determined to
characterize each of the different classes. The proposed patterns occurring in a specific tissue state. Salient patterns
method has been tested with plicated colon tumor for a specific state do not need to reveal similar behavior
data and the experimental results are evaluated in over all the same states. Instead, some