Classifiers based on rough mereology in analysis of DNA microarray data

There are numerous application of DNA microar-rays. The most frequently used application is the one used to research gene expression. The aim is, among others, to detect symptoms of illnesses in tissues, to predict the predisposition for some illnesses and for personal identification. The huge amount of information obtained from DNA microarrays in range of tens of thousands of genes causes many difficulties. Many algorithms, for instance, many rough set methods cannot be applied for this reason and due to low number of training objects tend to overfit. In this paper, we aim at presenting gene separation and classification methods by means of the granular classifier based on weighted voting, which was investigated recently by Polkowski and Artiemjew. Results of the research show that the obtained results are in many cases better than some standard methods such as rough set standard exhaustive classifier and k-nearest neighbor.