DPBC: Distance Based Possibilistic Biclustering With Application to Gene Ezpression Analysis
Biclustering is a key step in analyzing gene expression data by identifying patterns where subset of genes are co-related based on a subset of conditions. This paper proposes a new distance based possibilistic biclustering algorithm (DPBC), in which the average distances between rows and between columns of the bicluster are minimized and at the same time the size of the bicluster is maximized by computing the zeros of the derivative of appropriate objective function. The proposed algorithm uses the possibilistic clustering paradigm similar to another existing possibilistic biclustering algorithm PBC. Whereas PBC is based on residue our approach is applicable to any accepted definition for distances between pairs of rows or columns. Experimental study on the human dataset and several artificial datasets having different noise levels shows that the DPBC algorithm can offer substantial improvements over the previously proposed algorithms.
Possibilistic biclustering fuzzy clustering biclustering bidimensional clustering microarray data analysis
Mohamed A.Mahfouz Mohamed A.Ismail
Department of Computers and Systems Engineering Faculty of Engineering,Alexandria University Alexandria 21544,Egypt
国际会议
北京
英文
1-4
2009-06-11（万方平台首次上网日期，不代表论文的发表时间）