A GA-Based Classifier for Microarray Data Classification

This work presents an algorithm for generating the GA-based (Genetic Algorithm) classifier for microarray data classification. The microarray dataset comprises of a small number of samples with very high features. In order to construct the GA-based classifier, a number of informative features (genes) are selected. These features are divided into 2 groups (10 features or less in each group). The summation of gene expression values selected by GA in each group is then calculated and compared between groups. If the summation of the first group is greater than the other, it is classified as class 1; otherwise, it is classified as class 2. In the experiment, 3 microarray benchmark datasets for the 2-class problem are used. There are Lymphoma, Leukemia and Colon datasets. 10-Folds cross validation is used to test the performance of the proposed method. The experimental results show that the proposed GA-based classifier yields a good effectiveness in the 2-class microarray data classification comparing with the other methods.