New support vector machine-based approach over DNA chip data

DNA chip technology is an improved and highly scaled-up version of a 25 year old method to reveal very small changes in several hundreds or even thousands of genes in one step rather than searching one gene at a time. In this paper we used the outcome of this method to recognize the gene functional class cytoplasmic ribosomes, the cell cycle and DNA processing function by their gene expression level, with enforced support vector machines (SVM) by introducing new Kernel, specified for this problem.