Gene Expression Profile Classification Using Random Projection and Sparse Representation

A new gene expression profile classification scheme is developed in this study. Random projection is used for feature selection, and classification is formulated into a problem as finding sparse representations of test samples with respect to training samples. The sparse representation is computed by the l₁-regularized least square method. To investigate its performance, the proposed method is applied to three tumor gene expression datasets and compared with the combination of support vector machine (SVM) and two popular gene selection methods. The experimental results have shown that the performance of the proposed method is comparable with or better than those of SVM. In addition, the proposed method is more efficient than SVM as it has no need of model selection.