The Community for Technology Leaders
Green Image
Issue No. 05 - September/October (2011 vol. 8)
ISSN: 1545-5963
pp: 1273-1282
Chun-Hou Zheng , Qufu Normal University, Rizhao and The Hong Kong Polytechnic University, Hong Kong
Lei Zhang , The Hong Kong Polytechnic University, Hong Kong
To-Yee Ng , The Hong Kong Polytechnic University, Hong Kong
Simon C.K. Shiu , The Hong Kong Polytechnic University, Hong Kong
De-Shuang Huang , Tongi University, Shanghai
A reliable and accurate identification of the type of tumors is crucial to the proper treatment of cancers. In recent years, it has been shown that sparse representation (SR) by l_1-norm minimization is robust to noise, outliers and even incomplete measurements, and SR has been successfully used for classification. This paper presents a new SR-based method for tumor classification using gene expression data. A set of metasamples are extracted from the training samples, and then an input testing sample is represented as the linear combination of these metasamples by l_1-regularized least square method. Classification is achieved by using a discriminating function defined on the representation coefficients. Since l_1-norm minimization leads to a sparse solution, the proposed method is called metasample-based SR classification (MSRC). Extensive experiments on publicly available gene expression data sets show that MSRC is efficient for tumor classification, achieving higher accuracy than many existing representative schemes.
Tumors classification, sparse representation, metasample, gene expression data.

L. Zhang, S. C. Shiu, T. Ng, C. Zheng and D. Huang, "Metasample-Based Sparse Representation for Tumor Classification," in IEEE/ACM Transactions on Computational Biology and Bioinformatics, vol. 8, no. , pp. 1273-1282, 2011.
85 ms
(Ver 3.3 (11022016))