Hybrid feature selection methods for the Classification of Cancer in Micro-array Gene expression data: a Survey

For the study of disease diagnostics, microarray technology is widely used with gene expression rates.Researchers simultaneously studied the degree of expression of thousands of genes through the advent of DNA microarray technology.The Micro-array data analysis is the method to remove redundant and obsolete genes, to identify the most significant genes.Cancer identification is one of the most significant applications of the study of micro-array results.The efficiency of microarray technology depends on measurement precision, use of data processing techniques, research methods and statistical modeling.But still the curse of dimensionality and the curse of sparseness are a challenge to classify the gene expression profile.A collection of features (genes) is one of the most successful approaches to face these challenges.The function selection methods are used by eliminating noisy and nonrelevant features which develop classification performance to obtain an informative representation.Within the literature, there are numerous works to pick the main features from the microarray.This paper explores the new methods for selecting hybrid characteristics for selecting major genes from the findings of micro-array expression for the diagnosis of cancer.