In gene selection for cancer classification using microarray data, we define a mixed integer optimization model which is applied with the aim of selecting a small subset of cancer marker genes. Gene expression profiling has been successfully use to identity potential cancer diagnosis and therapeutic target in the past few years. Since there are often large number of potential genes, an important issue is to find a small subset of genes that are differentially expressed between different clinical outcomes (or) related subset to cancer patients survival time, and thus can be used to build prognosis and predictors. The study of...