IISA 2009 Paper

Many clustering techniques have been proposed for the analysis of gene expression data obtained from microarray experiments. However, choice of suitable method(s) for a given experimental dataset is not straightforward. KFCM algorithm has been widely applied in gene expression data analysis, but it is sensitive to the class center migration. Therefore, the fuzzy kernel clustering algorithm based on the kernel function and a dynamic weighted is proposed. The algorithm for the kernel function marking the introduction of the effectiveness function, missing data be filtered from the data sets and the original data is mapped to the high-dimensional feature space in each category of each gene and the right to assign a dynamic value, so that in the clustering algorithm iterative process of adaptive adjustment of weights, after analysis of the value of the size of the right to identify outliers, finally, getting the more ideal clustering result. Experiments showed that the proposed methods achieve better clustering effect than the fuzzy kernel clustering algorithm and possibility clustering algorithm, and is fast in convergence