Metasample Based Sparse Representation for Tumor Classification

Description

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 l1 -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 l1-regularized least square method. Classification is achieved by using a discriminating function defined on the representation coefficients. Since l1-norm minimization leads to a sparse solution, the proposed method is called metasample based SR classification (MSRC).
Extensive experiments on publicly available gene expression datasets show that MSRC is efficient for tumor classification, achieving higher accuracy than many existing representative schemes.

Nonlocally centralized sparse representation for image interpolation

Nonlocally centralized sparse representation for image denoising

https://en.wikipedia.org/wiki/Sparse_approximation