Tumor Classification via Sparse Representation Based on Metasample

In this paper, we propose a new method for tumor classification using gene expression data. The new method expresses each testing sample as a linear combination of a set of metasamples extracted from all the training samples. Classification is achieved by a defined discriminating functions using the coefficient vector for the metasamples extracted from each category, which is obtained by l <sub xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">1</sub> -regularized least square. Since l <sub xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">1</sub> -normminimization could leads to sparse solution, our approach can be named as metasample based sparse representation classification (MSRC). The experimental results show that our method is efficient for tumor classification.