The DNA microarray technology has capability to determine the levels of thousands of gene simultaneously in a single experiment. Analysis of gene expression is important in many fields of biological research in order to retrieve the required information. As time progresses, the illness in general and cancer in particular have become more and more complex and complicated, in detecting, analyzing and curing. We know cancer is deadly disease. Cancer research is one of the major area of research in medical field. Predicting precisely of different tumor types is a great challenge and providing accurate prediction will have great value in providing better treatment to the patients. To achieve this, data mining algorithms are important tools and the most extensively used approach to achieve important feature of gene expression data and plays an important role for gene classification. One of major challenges is to discover how to extract useful information from huge datasets. This paper presents recent advances in the machine learning based gene expression data analysis with different feature selection algorithms. Gene expression profiles, which represent the state of a cell at a molecular level, have great potential as a medical diagnosis tool. But compared to the number of genes involved, available training data sets generally have a fairly small sample size for classification. These training data limitations constitute a challenge to certain classification methodologies. Feature selection techniques can be used to extract the marker genes which influence the classification accuracy effectively by eliminating the un wanted noisy and redundant genes This paper presents a review of feature selection techniques that have been employed in micro array data based cancer classification and also the predominant role of SVM for cancer classification. © 2015 The Authors. Published by Elsevier B.V.