Disease-Related Gene Identification by an Improved Type-2 Fuzzy Method on Microarrays.

Type-2 fuzzy set can control the uncertainty information more effectively than conventional type-1 fuzzy set because its membership function is three-dimensional. In this paper we perform the identification of disease-related genes based on DNA microarray by type-2 fuzzy set theory and established type-2 fuzzy membership function to describe the differences of the gene expression values generated form normal people's genes and patient's. We also make a comparison between our type-2 fuzzy method with the traditional fuzzy method. The results suggest the type-2 FM-d values are significantly different, which make the identifications more reasonable and convincing than the original one. [ABSTRACT FROM AUTHOR]